PDA

View Full Version : Big Y vs Full Genomes



DillonResearcher
04-12-2016, 12:07 PM
I'd be interested to hear what peoples views are on this topic, DNA Day is coming up and if FTDNA do the same discount that they did last year I would be able to afford a Big Y test. I'm R-L2 with no matches and so it would be interesting to see what next gen sequencing would show.

I've been doing a bit of comparing of Full Genomes Y Elite with Big Y and it seems that the main difference (other than price) is that Y Elite covers more of the chromosome.

So my main question is what do you favour out of the two and why?

Mac von Frankfurt
04-12-2016, 12:36 PM
The current situation with Y-DNA testing is frustrating. One company refuses to upgrade their product while the other struggles with throughput. I went with FGC because the Y-DNA tree can only be defined to the fullest by the best available test at any given time. Also, I am on a very sparsely populated limb of the tree.

MitchellSince1893
04-12-2016, 12:57 PM
I ve ordered both. At the time there was a significant price difference between the two but now not so much. If I was ordering at this time I would go with fgc

RobertCasey
04-12-2016, 12:59 PM
I went with Full Genomes Corp before Big Y was announced. I am the only L226 tester out of 48 NGS testers that used FGC vs. Big Y. I have 12 private YSNPs that Big Y does not test, so there is a real difference in YDNA coverage. I think there needs to be a small mixture of higher coverage FGC tests mixed in with primarily Big Y tests. I think it should be around 5 to 10 % of the NGS tests need to be FGC tests to uncover more older major branches. However, these FGC tests should not be random. They need to be either be very close to each other (to maximize private YSNPs that define branches within a surname cluster) or near the trunk of the tree (the bottleneck where most of the private YSNPs are stuck at).

With L226, around one third of the private YSNPs are stuck with no new major branches other than my original FGC test - FGC5660 > FGC5628 > FGC5659 > FGC5639. The only downside of FGC, they can be painfully slow some times but will provide around 10 to 20 % more private YSNPs to test than Big Y which worth the wait for the more serious testers. Also, you need to have a strategy of testing your private YSNPs as well. Just waiting for more Big Y results to reveal branches is very costly and rarely reveal genealogical time frame YSNPs.

MacUalraig
04-12-2016, 02:01 PM
I've tested twice with FGC now (Y Elite and WGS), it took longer but the wait was worth it - my branch defining SNPs aren't even on the BigY.

Clinton P
04-12-2016, 04:04 PM
A clue to the value of these products is in their names.... Big Y and Full Genomes.

It implies (and quite rightly so) that Big < Full.

I have done both. Full Genomes offers quantity and quality but at the expense of time taken, whereas Big Y costs less and is relatively quick but has less coverage of the Y chromosome.

You'll certainly get more SNPs/$ with Full Genomes.

Clinton P

DillonResearcher
04-12-2016, 06:52 PM
Great, thanks for all the replies, so it looks like saving for a bit longer for Full Genomes is better in the long run.

VinceT
04-12-2016, 06:54 PM
I was one of the first Y-Elite 1.0 participants, and have had no regrets, other than the price I had paid compared with the price Full Genomes currently offers for Y-Elite 2.1.

I ordered Big Y for my maternal uncle, only because I wasn't able to afford another Y-Elite test, and I was too lazy to arrange the collection of a third DNA sample from him. (The first two were for 23andMe and FTDNA respectively). If I could afford it, I would have ordered the 30x WGS from Full Genomes for everyone. But on the same note, I'd prefer read lengths that exceed 500 or even 10,000 bases, so it's perhaps best to put it off.

Right now, Big Y uses read lengths between 150 to 165 nucleotides, and Y-Elite 2.x is well over 200, striving towards 250. Longer read lengths, along with increased depth of coverage, help significantly with increasing the number and accuracy of STRs that can be analyzed.

Y-Elite is better than Big-Y in nearly every respect except price and processing time.

TigerMW
04-13-2016, 01:14 PM
...
I ordered Big Y for my maternal uncle, only because I wasn't able to afford another Y-Elite test, and I was too lazy to arrange the collection of a third DNA sample from him..
This is the convenience factor.

So beyond except price and processing time you must add convenience.

Y-Elite appears to very good, so please don't take this as a dis on Y-Elite, but Big Y has something else, or rather FTDNA does.

FTDNA has a large customer set with a fairly comprehensive product line and DNA storage. This provides two other advantages for Big Y, besides the convenience advantage and you get a stronger matching database.

General academic research of the the ancient branching is one thing, but for most of us, it does not matter what test you take if other people related to you don't get measured at the same places in the same methods. Genetic genealogy is built on the relative comparison method. There is no getting around this. This neuters the Y-Elite coverage advantages for the average consumers.

It is key to plan out how you will get others to test who are potentially related to you. It is a team sport.

Actually, there is another advantage for Big Y beyond the four cited here. That advantage is an early feeder of SNPs into a broader set of consumer SNP packs/panels. I think we all want to get our youthful SNPs tested by as many people as possible.

More to add on the pro's and con's is the company factor. FTDNA has stood the test of time and competition. They are accessible in courts and have something to lose. They won't merely disband and go bankrupt if faced with a problem. FTDNA is most likely to be here tomorrow. Others may not. What are the succession plans, etc.? A con is FTDNA service is not personal, that's for sure. Their customer communications are poor.

Summary of Big Y advantages.
1. Lower price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

Clinton P
04-13-2016, 01:59 PM
This is the convenience factor.

FTDNA has a large customer set with a fairly comprehensive product line and DNA storage. This provides two other advantages for Big Y, besides the convenience advantage and you get a stronger matching database.




Surely you mean stronger mis-matching database. ;)

Clinton P

TigerMW
04-13-2016, 02:15 PM
Surely you mean stronger mis-matching database. ;)

Clinton P
You are probably speaking of the Big Y matching web tool. That [email protected]#$'s although it can be manipulated into some use. In some subclades, this is used extensively for recruiting new people to projects.

Regardless, there is a very large group of Big Y testers with tests already completed. We have over 1,500 for just R1b-L21 alone. As long as we can get folks to join projects, we know about these people and how to contact them. We can also cross-check with Y STR matching.

RobertCasey
04-13-2016, 02:17 PM
Summary of Big Y advantages.
1. Lower price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

I really like the expansion of your list of issues involved. However, I think you are getting a little too FTDNA biased these days as well.

1. Lower price - this use to be a major issue - but now the prices are not that far apart considering the higher coverage & better analysis. But it does remain a factor since it does cost less per test.
2. Faster test processing time - this is a major advantage of FTDNA - but when you keep your product static with no improvements in quality or price, it is much easier doing very little to improve your product. But the recent delay in Big Y downloads is a recent bump in the road for FTDNA as well.
3. Convenience - there is some truth here. It is probably easier to get people to test with FTDNA. But this comes at a cost as well. Since FTDNA is the market leader - they are much slower to improve their products. Also, they are virtually killing individual YSNP testing for testing private YSNPs. The cost per YSNP discovered by Big Y for L226 is $1,400 per branch. Testing private YSNPs is currently only $140 per branch (but are smaller branches).
4. Database - since almost all analysis is being done via BAM files and FTDNA does not post Big Y results into their YSNP reports, this normal advantage of a common database is not really an advantage.
5. Big Y does feed into their haplotree development and YSNP packs which is a major advantage. But their resistance to offer individual YSNPs also really impedes growth of the haplotree as well. Without comprehensive individual YSNP testing, we are forced to shell out a lot more funds in the long run.
6. Without competition, FTDNA will be less innovative which will slow progress of technology and drive up overall costs in the long run. Big Y was a direct response to FGC's offering (would you still accept the pace of discovery of Walk the Y). Because they do not offer individual YSNP testing like YSEQ does, the cost per branch discovered remains the same as Walk the Y (around $1,400 per branch). Without YSEQ's YSNP panel tests, SNP packs would be much slower to come about. Also, FTDNA's approach to SNP panel tests is more profitable and can test more YSNPs but it requires more setup costs. The SNP packs are sadly out of date and are not updated as often - which costs the genetic community more in the long run.

Do not get me wrong, FTDNA is a great company and deserves to be the dominant leader in YDNA. However, the trend to be reactive only to competition is quite disturbing. Their trend to not test for individual YSNPs and only offer Big Y also shows that this company is making a lot of profit at the expense of progress and overall costs to the genetic community. We should strive for a 5 to 10 % mix of FGC NGS tests in with all the Big Ys, as this will discover more major branches that Big Y will never reveal and will also reveal some very nice genealogical YSNPs as well.

TigerMW
04-13-2016, 03:23 PM
I really like the expansion of your list of issues involved. However, I think you are getting a little too FTDNA biased these days as well.

Thank you for you opinion. Please note I did not call anyone biased here or any way criticize some poster. I'm just making arguments in logic, or at least I'm trying to. :)



1. Lower price - this use to be a major issue - but now the prices are not that far apart considering the higher coverage & better analysis. But it does remain a factor since it does cost less per test.

Agreed, generally, but I still think prices are far the most important issue and no one has it right (for the market). We need $299 and $199 priced tests. Then we will really get somewhere. NGS testing is for the elite. BTW, this ties to another reply you made on another thread that does not reconcile with what Richard R and Vince T are finding from FGC and TK - that the NGS "enrichment cost" is high, like $300-$400, so a test like Big Y or Y Elite can not come down in price that much (and is not that profitable.)
These NGS tests really are for the elite customers. I call them lead explorers.



2. Faster test processing time - this is a major advantage of FTDNA - but when you keep your product static with no improvements in quality or price, it is much easier doing very little to improve your product. But the recent delay in Big Y downloads is a recent bump in the road for FTDNA as well.

The temporary suspension is over so that is bit moot point now. However, you could still have gotten a lot done with VCF files anyway - witness Big Tree. I'm not dissing BAM files. They are important and I have 2nd interpretation on my own in the hopper.


3. Convenience - there is some truth here. It is probably easier to get people to test with FTDNA. But this comes at a cost as well. Since FTDNA is the market leader - they are much slower to improve their products. Also, they are virtually killing individual YSNP testing for testing private YSNPs. The cost per YSNP discovered by Big Y for L226 is $1,400 per branch. Testing private YSNPs is currently only $140 per branch (but are smaller branches).
BTW, did you see my personal email to you on L226? There are those in pursuit of this in surname projects. In full disclosure, I'm not a big fan of individual SNP testing any more. It feels like a wild goose chase sometimes and I hate to be on the receiving end of testing for other peoples' SNPs. You may not like my comment here, but at this level of resolution people really should be doing NGS testing, or living with the vagaries of STRs. Anything else is, well...


4. Database - since almost all analysis is being done via BAM files and FTDNA does not post Big Y results into their YSNP reports, this normal advantage of a common database is not really an advantage.Robert, FTDNA started reporting Big Y results in their Y DNA SNP public project screen last year. They try to update known SNPs that are derived and negative results downstream but all of this depends on their haplotree. Their haplotree is changing dramatically (and painfully) every week. I download Y DNA public project screens about every two weeks. Most Big Y testers have a change from the prior update. This indicative of the changes to the haplotree.

We do have admins in some projects who go on-line every time a new Big Y result comes in and check the matching. They do find people who "we didn't know about."


5. Big Y does feed into their haplotree development and YSNP packs which is a major advantage. But their resistance to offer individual YSNPs also really impedes growth of the haplotree as well. Without comprehensive individual YSNP testing, we are forced to shell out a lot more funds in the long run.
I agree, their hiatus on new SNP additions to their Advanced Tests individual offering is not good. Their stated goal is to have everything in an SNP Pack also available as a individual SNP test. They are not there yet, surely. It is my guess they've spent the first quarter scrubbing their SNPs and will move forward with additions as they finish their process renovation.


6. Without competition, FTDNA will be less innovative...
I absolutely agree, competition is good, very good. FGC and YSEQ have been a great service to the marketplace. Kudos to them.

This is not you but I've had advocates suggest I "throw business" in those directions to help balance the competition, and also just because they need financial support (families to feed). I understand the empathy, but I just can't do that. To me, it is up to the individual consumer what they choose. A good open discussion helps so seeking advice publicly is always a good thing.

As an aside, I think I've personally helped drive prices down. I try to not brag but I've seen the price drops where certain events cause a reaction. I think that is good for the consumer.

Mac von Frankfurt
04-13-2016, 03:28 PM
It is often said that NGS is for experienced testers only. I do not think this is true anymore. For the new tester what could be more convenient than sending one sample to one lab and getting one test and then sharing the results with YFull or The Big Tree or a Yahoo Group? And overall it is very economical in both time and money compared to taking one test after another and never knowing exactly where you are or when it will end.

So don’t be afraid to be a lead dog and enjoy the view. Take the best test available.

I understand Project Administrators have to deal with the current reality and the history of Y-DNA testing but in my perfect world there would only be two types of tests: the deepest test available for leaders, and complete panels based on those deep tests for followers. If both of these tests were provided by the same company (which they are not currently) the deep tests could actually be subsidized by the panels as loss leaders. Just imagine how that would catalyze the growth of the Y tree. Yes it is necessary and right to bring the people who took tests in the old paradigm along and I took 67 STRs at FTDNA. But why should new testers have to go through all of that when there is a much simpler alternative for them that is also extremely valuable to the overall effort?

FGC Corp
04-13-2016, 04:26 PM
Big Y was a direct response to FGC's offering (would you still accept the pace of discovery of Walk the Y). Because they do not offer individual YSNP testing like YSEQ does, the cost per branch discovered remains the same as Walk the Y (around $1,400 per branch). Without YSEQ's YSNP panel tests, SNP packs would be much slower to come about. Also, FTDNA's approach to SNP panel tests is more profitable and can test more YSNPs but it requires more setup costs. The SNP packs are sadly out of date and are not updated as often - which costs the genetic community more in the long run.

Do not get me wrong, FTDNA is a great company and deserves to be the dominant leader in YDNA. However, the trend to be reactive only to competition is quite disturbing. Their trend to not test for individual YSNPs and only offer Big Y also shows that this company is making a lot of profit at the expense of progress and overall costs to the genetic community. We should strive for a 5 to 10 % mix of FGC NGS tests in with all the Big Ys, as this will discover more major branches that Big Y will never reveal and will also reveal some very nice genealogical YSNPs as well.

Actually, I started doing this precisely because it wasn't available, and not initially, with the idea of making this a large enterprise. I even advised a certain person a few years ago that we were going to do it. The second point, is that I took a different approach from the beginning which is that we wanted to focus 100% on quality, since it would be challenging to compete on customer base or volume.

Even in the case of the whole genome product, what we do isn't available as an ancestry product from George Church's Veritas company. Secondly, your data doesn't go to the US medical system. I wanted a pathway personally so I could order NGS tests for myself to begin with. We did it.

My focus on quality has been vindicated since we were selected twice for a non-bid contract and solved the problem twice with 100% success. That work has never been done before.

Essentially, I can order (or anyone else), can order any existing NGS test. We don't provide clinical reports, but the customer has the widest possible access to a variety of technologies.

We have a different focus. To the extent that someone else is already providing a certain type of test, we probably won't do it.Moving forward, we're interested in doing things that aren't being done or which other companies aren't willing to do.

For example, the single cell sequencing service is for researchers only. But we can provide services to the major US research institutions, such as Salk, Univ Virginia, etc.

No one else can match us there. They can't do the single cell sequencing with the same level of accuracy.


Actually, I started doing this precisely because it wasn't available, and not initially, with the idea of making this a large enterprise.

Actually, I also wanted to prove a point (not to the community) but to a couple people I know personally. That was done too, I think.

DillonResearcher
04-13-2016, 04:37 PM
This is the convenience factor.

So beyond except price and processing time you must add convenience.

Y-Elite appears to very good, so please don't take this as a dis on Y-Elite, but Big Y has something else, or rather FTDNA does.

FTDNA has a large customer set with a fairly comprehensive product line and DNA storage. This provides two other advantages for Big Y, besides the convenience advantage and you get a stronger matching database.

General academic research of the the ancient branching is one thing, but for most of us, it does not matter what test you take if other people related to you don't get measured at the same places in the same methods. Genetic genealogy is built on the relative comparison method. There is no getting around this. This neuters the Y-Elite coverage advantages for the average consumers.

It is key to plan out how you will get others to test who are potentially related to you. It is a team sport.

Actually, there is another advantage for Big Y beyond the four cited here. That advantage is an early feeder of SNPs into a broader set of consumer SNP packs/panels. I think we all want to get our youthful SNPs tested by as many people as possible.

More to add on the pro's and con's is the company factor. FTDNA has stood the test of time and competition. They are accessible in courts and have something to lose. They won't merely disband and go bankrupt if faced with a problem. FTDNA is most likely to be here tomorrow. Others may not. What are the succession plans, etc.? A con is FTDNA service is not personal, that's for sure. Their customer communications are poor.

Summary of Big Y advantages.
1. Lower price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

Thanks, that was very interesting to get a more positive view of the Big Y.

Having weighed up what's been said I think that if FTDNA does put it on sale on DNA Day I'll go with the Big Y since although it's not a comprehensive as Y Elite it's the best I can afford at the moment (not likely to change for a while) and the positives given above have given me a bit of confidence in Big Y.

FGC Corp
04-13-2016, 04:41 PM
Thanks, that was very interesting to get a more positive view of the Big Y.

Having weighed up what's been said I think that if FTDNA does put it on sale on DNA Day I'll go with the Big Y since although it's not a comprehensive as Y Elite it's the best I can afford at the moment (not likely to change for a while) and the positives given above have given me a bit of confidence in Big Y.

As far as payments go, we offer flexible payment plans. I might as well respond here and clarify. [I'll pass on Big Y vs Y Elite comparisons, etc]

Edit:
There are a couple other places where we discuss various other things we're doing:

1. Facebook:
https://www.facebook.com/login/?next=https%3A%2F%2Fwww.facebook.com%2Fgroups%2Ffu llgenomesY%2F

2. Google group:
https://groups.google.com/forum/?hl=en#!forum/y-chromosome-fgc-analytical-discussions

#2 We discuss more advanced topics at the google group.

TigerMW
04-13-2016, 04:49 PM
It is often said that NGS is for experienced testers only. I do not think this is true anymore. ...
I agree that it is not required for a tester to feel experienced. Sometimes they have to be experienced to understand what you get with different kinds of testing, though.

This is different than experienced, when I posted below, "These NGS tests really are for the elite customers. I call them lead explorers."

It takes desire and money to be a lead explorer. Experience is a good thing, but is not required. There is help.

If you guys keep me talking, I'll pull out my Lewis and Clark Expedition analogy.:)

Amerijoe
04-13-2016, 04:56 PM
Having been involved in non-bio testing procedures, quality of the product was of the upmost concern. I've researched both companies and decided FGC in my opinion and other members as well has better quality control. I am aware of past scheduling issues, but decided quality trumps promptness. Due to the rarity of Y15121 members I also wanted to have as much information extracted as possible, again FGC is presented as offering more. Your decision should be weighed with the same process as with any important purchase.

VinceT
04-13-2016, 05:23 PM
Surely you mean stronger mis-matching database. ;)

Clinton P

Indeed!

FTDNA's matching/mismatching algorithms are pretty damned feeble. They show "matches" for a R-U106>FGC396>L199 guy with men under R-P311*, R-P312>DF27>Z31644>FGC13109 and R-P312>DF27>Z34609>Z2571>FGC11388>Y17787, but not with men under R-U106>A2150 or R-U106>S18632. It all depends on what variants they have bothered to add to their database. If two Big Y's mismatch on 40 variants, but only two are recognized by FTDNA, they show up as a match. If they mismatch on only six variants, of which five are recognized by FTDNA, they won't show up as a match because FTDNA caps mismatches at four "known" variants.

kinman
04-14-2016, 02:44 AM
I personally went with Big Y as a stepping stone, not only because it was convenient (no need for another kit, and much faster), but a good discount price at just the right time. Although I didn't get as many SNPs, I probably got about as many SNPs per dollar spent. And as has been hinted at, having more SNPs than anyone else in your haplogroup might feel nice at first, but it can probably be frustrating when you are left waiting for others to catch up (enough to make meaningful comparisons).
Therefore, I think an important factor is how many others in your haplogroup are getting tested and what tests they can afford. Being on the cutting edge is nice if you can afford it, but a majority of us can't. And frankly most of my Big Y "matches" in the past few months haven't been anywhere near my haplogroup anyway, so I wouldn't be any better off with Y Elite at this time. Therefore, I am waiting for prices on full genomes to fall a lot more, and then decide how much coverage is optimal for the prices at that time (10X. 30X, 50X, or more). Until then, I am just discovering just how much I can find out about my other lines in my latest test (Family Finder). Gives me plenty to do while waiting for others to catch up on my all male lineage (which is just a tiny fraction of what I want to explore).

Mac von Frankfurt
04-14-2016, 03:24 AM
I personally went with Big Y as a stepping stone, not only because it was convenient (no need for another kit, and much faster), but a good discount price at just the right time. Although I didn't get as many SNPs, I probably got about as many SNPs per dollar spent. And as has been hinted at, having more SNPs than anyone else in your haplogroup might feel nice at first, but it can probably be frustrating when you are left waiting for others to catch up (enough to make meaningful comparisons).
Therefore, I think an important factor is how many others in your haplogroup are getting tested and what tests they can afford. Being on the cutting edge is nice if you can afford it, but a majority of us can't. And frankly most of my Big Y "matches" in the past few months haven't been anywhere near my haplogroup anyway, so I wouldn't be any better off with Y Elite at this time. Therefore, I am waiting for prices on full genomes to fall a lot more, and then decide how much coverage is optimal for the prices at that time (10X. 30X, 50X, or more). Until then, I am just discovering just how much I can find out about my other lines in my latest test (Family Finder). Gives me plenty to do while waiting for others to catch up on my all male lineage (which is just a tiny fraction of what I want to explore).

How can a Big Y match not be exactly your haplogroup, by definition? I understand an STR predicted match not panning out but a NGS match should share a terminal SNP with you. Maybe my DNA vocabulary is failing me.

GailT
04-14-2016, 04:27 AM
How can a Big Y match not be exactly your haplogroup, by definition? I understand an STR predicted match not panning out but a NGS match should share a terminal SNP with you. Maybe my DNA vocabulary is failing me.

I think the problem is in the way that FTDNA identifies matches. Last time I checked most of my BigY "matches" were not in my subclade R1b-L2, so obviously they were not really matches.

Also, when I tested BigY it did not have very good coverage of the y genome. I tested both the BigY and FGC FullY for my R1b-L2 paternal line. BigY found 25 private SNPs compared to 50 private SNPs in the FGC FullY test. All 25 of the BigY SNPs were also included in the FullY results.

RobertCasey
04-14-2016, 04:44 AM
BTW, did you see my personal email to you on L226? There are those in pursuit of this in surname projects. In full disclosure, I'm not a big fan of individual SNP testing any more. It feels like a wild goose chase sometimes and I hate to be on the receiving end of testing for other peoples' SNPs. You may not like my comment here, but at this level of resolution people really should be doing NGS testing, or living with the vagaries of STRs. Anything else is, well...

I did not see any email from you (however, I just discovered that all gmail.com was being blocked which I just fixed). I remain a big fan of individual testing of YSNPs for three primary reasons: 1) they discover more recent mutations - under L226, we now have found two Casey only YSNP mutations (which covers both known Casey surname clusters); 2) If you pay attention to the signature associated with the NGS testers, you can find several good testing candidates for testing YSNPs (signatures are still useful even at this level); 3) The cost per new branch found is around 10 % the costs of yet more Big Y tests. However, these are much smaller scope branches (which is my main area of interest since that is why I am doing this as a hard core genealogist first). I only help others with older branches as we have to inch our down to more recent YSNPs.

As far as competition for NGS testing, I only advocate 5 to 10 percent of NGS tests should be Full Genomes for two purposes only: 1) if you are interested in finding genealogical YSNPs - two FGC tests that are close YSTR matches will reveal more private YSNPs which could be even younger branches; 2) With L226, there are around 30 % of private YSNPs stuck in the main trunk of the L226 haplotree. Using higher resolution tests just below the last major branch could reveal the most significant branches to break up the trunk bottleneck. After the first L226 NGS test (my test using FGC), 48 Big Y tests later, no new branches between these major branches. I still have 12 private YSNPs not tested by Big Y tests. If the original ratios of types of YNSPs track the ones that are tested, there should be mostly more L226 equivalents - but will also include one major branch and 2 or 3 new private YSNPs as well. I am very disappointed that so many private YSNPs remain private (which does not make sense to me). But there is almost always one branch and sometimes two branches that can be found with individual YSNP testing. The branches usually only cover five to ten submissions, so they are quite small branches (they way I like them).

Lugus
04-14-2016, 05:00 AM
I think the problem is in the way that FTDNA identifies matches. Last time I checked most of my BigY "matches" were not in my subclade R1b-L2, so obviously they were not really matches.

Indeed I see your name in the list of my Big Y matches. In a way they are real matches, just very (very) remote. It's not by chance that my first non-DF27 matches are U152. After all we're P312 and share another SNP below, but I also have some U106 matches. It all the depends on your perspective. I understand most Americans are looking for recent relatives but I'm more interested in deep ancestry and especially to know more about what our ancestors were up to in the remote past.

FGC Corp
04-15-2016, 09:44 PM
This is the convenience factor.

So beyond except price and processing time you must add convenience.

Y-Elite appears to very good, so please don't take this as a dis on Y-Elite, but Big Y has something else, or rather FTDNA does.

FTDNA has a large customer set with a fairly comprehensive product line and DNA storage. This provides two other advantages for Big Y, besides the convenience advantage and you get a stronger matching database.

General academic research of the the ancient branching is one thing, but for most of us, it does not matter what test you take if other people related to you don't get measured at the same places in the same methods. Genetic genealogy is built on the relative comparison method. There is no getting around this. This neuters the Y-Elite coverage advantages for the average consumers.

It is key to plan out how you will get others to test who are potentially related to you. It is a team sport.

Actually, there is another advantage for Big Y beyond the four cited here. That advantage is an early feeder of SNPs into a broader set of consumer SNP packs/panels. I think we all want to get our youthful SNPs tested by as many people as possible.

More to add on the pro's and con's is the company factor. FTDNA has stood the test of time and competition. They are accessible in courts and have something to lose. They won't merely disband and go bankrupt if faced with a problem. FTDNA is most likely to be here tomorrow. Others may not. What are the succession plans, etc.? A con is FTDNA service is not personal, that's for sure. Their customer communications are poor.

Summary of Big Y advantages.
1. Lower price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

One more item:

Depending on the estimate, between 15,000 and 20,000 SNPs would not yet have been discovered if our work had not been done, because they are not covered on any other DTC Y chromosome test.

TigerMW
04-15-2016, 09:56 PM
.. Depending on the estimate, between 15,000 and 20,000 SNPs would not yet have been discovered if our work had not been done, because they are not covered on any other DTC Y chromosome test.
Agreed. FGC has been an irreplaceable research engine. Thank you.

Kwheaton
04-15-2016, 10:18 PM
I see value in both---

The analysis from Big Y is not there---it all depends on Project Administrators and many of the so called "novel SNPS" are misleading...but the comparison aspects are nice if you are able to red the tea leaves....

Still if you are going to spend that amount of money I would go with the YElite if there is no one else matching you who has done a NGS test. If someone already has then its a matter of your goals and budget. I do not think there is a one size fits all answer. I have used a combination of FTDNA, YSEQ and Full Genomes products.

FGC Corp
04-19-2016, 03:11 PM
One more point:

For certain types of projects, researchers need our level of resolution and coverage to consistently distinguish between recent family lines, particularly when the common ancestor was born after 1750 for example. There's a clear example in R1b-CTS2509 of that. [I happen to be under R1b-CTS2509 myself so I know the person in question from a few years back].

As for price, price scales directly with the size of the test. If our test had the same coverage the price would be close to FTDNA's. One pays more for more data, it's that simple.

RobertCasey
04-19-2016, 06:21 PM
For certain types of projects, researchers need our level of resolution and coverage to consistently distinguish between recent family lines, particularly when the common ancestor was born after 1750 for example. There's a clear example in R1b-CTS2509 of that. [I happen to be under R1b-CTS2509 myself so I know the person in question from a few years back].

I 100 % agree with this issue, the more private YSNPs that are revealed with higher coverage NGS tests, the more likely you are to find more recent YSNPs that are genealogical in nature.

The other major advantage of higher coverage, at least for R-L226, higher coverage also will discover more major branches. I believe that between 5 and 10 % of NGS tests should be Y Elites to discover these major branches. To date, L226 has a major trunk of the haplotree that was defined from the first NGS test from Y Elite. This trunk is FGC5660>FGC5628>FGC5659. Just around 1/3 of the private YSNPs are stuck in this trunk with no other branches found. These are FGC5660*, FGC5628* and FGC5659* (where * means no other downstream branches were positive). We have 24 branches yet 1/3 of private YSNPs are stuck under three branches. If my 12 private YSNPs (not tested by Big Y) track the distribution of those tested by Big Y tests, 8 would be yet more L226 equivalents, 1 would be a new major branch and 3 would be new private YSNPs that could be more recent genealogical branches. So it is not only more recent YSNPs as an advantage, but also more significant branches would be discovered as well. These higher coverage tests should not be random, they should be from parts of the haplotree that have these huge bottlenecks of private YSNPs. For me, FGC5660*, FGC5628* and FGC5659* individuals should test with Y Elite to reveal major new branches where over 100 high quality private YSNPs are stuck at.

FGC Corp
04-19-2016, 06:50 PM
I 100 % agree with this issue, the more private YSNPs that are revealed with higher coverage NGS tests, the more likely you are to find more recent YSNPs that are genealogical in nature.

The other major advantage of higher coverage, at least for R-L226, higher coverage also will discover more major branches. I believe that between 5 and 10 % of NGS tests should be Y Elites to discover these major branches. To date, L226 has a major trunk of the haplotree that was defined from the first NGS test from Y Elite. This trunk is FGC5660>FGC5628>FGC5659. Just around 1/3 of the private YSNPs are stuck in this trunk with no other branches found. These are FGC5660*, FGC5628* and FGC5659* (where * means no other downstream branches were positive). We have 24 branches yet 1/3 of private YSNPs are stuck under three branches. If my 12 private YSNPs (not tested by Big Y) track the distribution of those tested by Big Y tests, 8 would be yet more L226 equivalents, 1 would be a new major branch and 3 would be new private YSNPs that could be more recent genealogical branches. So it is not only more recent YSNPs as an advantage, but also more significant branches would be discovered as well. These higher coverage tests should not be random, they should be from parts of the haplotree that have these huge bottlenecks of private YSNPs. For me, FGC5660*, FGC5628* and FGC5659* individuals should test with Y Elite to reveal major new branches where over 100 high quality private YSNPs are stuck at.

The resolution can be even better than we report, because a number of ** SNPs have been subsequently validated by YSEQ as real. Secondly, the RM-STRs can be used to further separate out lines. Clearly, this would add to the overall cost to do additional validation, but one can get very close to 1 marker per generation using those additional approaches.

FGC Corp
04-19-2016, 08:21 PM
I'd also like to comment on the processing issue. We're catching up after our roadblock last year:

FGC has delivered 196 Y Elite results since August 1, 2015.

Batches 9004,9005,9006,and 9007 have been delivered (Y Elite 2.0)
Pilot Batch 8002 (Y Elite 2.1 has been delivered)

132 samples are sequencing now.

JamesKane
04-19-2016, 11:32 PM
I tend to like numbers when comparing similar products. The stats use the new b38 reference, so there are small variations with what is reported by the vendors.

Big Y - 9,284,386 callable loci (over 318 tests) = 15,604 bp / $
Y Elite 2.0 - 14,377,654 callable loci (over 14 tests) = 18,315 bp / $

The increased coverage proportionally increases the number of variants detected, which helps to break up blocks with common ancestors after 1000-1500 AD. There is also the added benefit in helping discover variants like Z39589.

That said when the sale windows reduce Big Y pricing, the value situation does reverse. This comes with the cost of losing the opportunity to get the extra 5,000,000 bases sequenced economically.

While most of the discussion will be about these two tests, we also need to keep the WGS options in mind. The 30x WGS sample I have access to has coverage on par with the best of the Y Elite 2.0s. The 15x GenomeGuide will outperform Big Y at least while giving the rest of the DNA results and potential to take up to 30x later. This is actually the route I would have went had my 1.0 not been in process when the WGS was introduced.

ChrisR
04-20-2016, 09:05 AM
new b38 reference
Big Y - 9,284,386 callable loci (over 318 tests) = 15,604 bp / $
Y Elite 2.0 - 14,377,654 callable loci (over 14 tests) = 18,315 bp / $
Thanks, your analysis and stats are very useful. Y Elite 1 had very stable coverage while BigY more a "Big Variance". Could you calculate for example the 95% Confidence interval of the callable loci standard deviation (Margin of error)?
I would be interested also in the coverage area (bp ≥1Ũ coverage).

FGC Corp
04-20-2016, 06:20 PM
I tend to like numbers when comparing similar products. The stats use the new b38 reference, so there are small variations with what is reported by the vendors.

Big Y - 9,284,386 callable loci (over 318 tests) = 15,604 bp / $
Y Elite 2.0 - 14,377,654 callable loci (over 14 tests) = 18,315 bp / $

The increased coverage proportionally increases the number of variants detected, which helps to break up blocks with common ancestors after 1000-1500 AD. There is also the added benefit in helping discover variants like Z39589.

That said when the sale windows reduce Big Y pricing, the value situation does reverse. This comes with the cost of losing the opportunity to get the extra 5,000,000 bases sequenced economically.



We were selected for a research project precisely because there is no readily available commercial alternative for what they needed, ie maximum possible coverage [given the current limits of the best sequencing technology available]. I can't discuss the project further because of a NDA.

JamesKane
04-20-2016, 11:54 PM
Thanks, your analysis and stats are very useful. Y Elite 1 had very stable coverage while BigY more a "Big Variance". Could you calculate for example the 95% Confidence interval of the callable loci standard deviation (Margin of error)?
I would be interested also in the coverage area (bp ≥1Ũ coverage).

There is a dynamic version of what I had previously published in my blog in the works: http://www.it2kane.org/2015/11/ngs-y-chromosome-alignment-on-grch38/ This post used ≥ 1x coverage values.

The new page will only include Y stats, unlike the older methodology. I'll incorporate the standard deviations in the work in progress.

TigerMW
04-21-2016, 01:52 PM
I think it is fait accompli for the average genetic genealogy tester. The Big Y train has left the station.
The Big Y advantages I summarized still apply.

1. Lower absolute entry price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

Now is a great time to jump on board as the FTDNA has their Big Y priced at $460 for DNA Day sales (April 21-26). Unless FTDNA has some magic, according to reports from FGC and TK the high enrichment costs of NGS testing mean FTDNA must hardly be making money on this. BTW, I'm a bit skeptical on how much the enrichment costs are but if they are that much, this is really a good deal.

The momentum of the train is what's important, though. Big Y orders in my projects trickle in nearly every week without special incentives. They just happen. This means critical mass in the market has been reached. These promotion events like the $460 just add an additional carload to the train.

Since genetic genealogy is only about comparisons of test results between people, the growth in the Big Y database is critical. We will find people and surprises we didn't know about.

I think FGC Elite looks like a great test, but even if you do it (and assuming you have that kind of money) you should consider getting Big Y anyway to get registered in the Big Y database and find people as they find you. Many people have done both tests. It may seem crazy but if you are avid and can afford it, it's worth it. These folks are just true pioneers.

A fellow poster John estimated the Big Y penetration based on some major R1b haplogroup project estimations I've done. His numbers are rough but he had L21 at 40% Big Y penetration of the total L21+ confirmed people [EDIT: I recalculated and think it should be 30% for L21]. He had L238 at 60%. He had U106 at 25%. I've got DF19 at 69%. I think DF19 is the winner with 145 Big Y's out of 210 people.

The net is there getting up to several thousand people in R1b alone with Big Y results in their database.



This is the convenience factor.

So beyond except price and processing time you must add convenience.

Y-Elite appears to very good, so please don't take this as a dis on Y-Elite, but Big Y has something else, or rather FTDNA does.

FTDNA has a large customer set with a fairly comprehensive product line and DNA storage. This provides two other advantages for Big Y, besides the convenience advantage and you get a stronger matching database.

General academic research of the the ancient branching is one thing, but for most of us, it does not matter what test you take if other people related to you don't get measured at the same places in the same methods. Genetic genealogy is built on the relative comparison method. There is no getting around this. This neuters the Y-Elite coverage advantages for the average consumers.

It is key to plan out how you will get others to test who are potentially related to you. It is a team sport.

Actually, there is another advantage for Big Y beyond the four cited here. That advantage is an early feeder of SNPs into a broader set of consumer SNP packs/panels. I think we all want to get our youthful SNPs tested by as many people as possible.

More to add on the pro's and con's is the company factor. FTDNA has stood the test of time and competition. They are accessible in courts and have something to lose. They won't merely disband and go bankrupt if faced with a problem. FTDNA is most likely to be here tomorrow. Others may not. What are the succession plans, etc.? A con is FTDNA service is not personal, that's for sure. Their customer communications are poor.

Summary of Big Y advantages.
1. Lower price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

FGC Corp
04-21-2016, 02:13 PM
I think it is fait accompli for the average genetic genealogy tester. The Big Y train has left the station.
The Big Y advantages I summarized still apply.
.

Not for very prestigious customers I work with Mike. FTDNA is inadequate for their purposes. They can't achieve their goals at all with that limited data set.

No, not a fait accompli. Bunk. In fact, our sales have increased dramatically for a few research projects recently, precisely because of the advantages already enumerated previously by myself and by other posters in this thread.

They're not in your project, so you're not seeing it.

kinman
04-21-2016, 02:27 PM
I heard that Big Y is going to soon be on sale for about $460 to celebrate DNA day.
Will there also be a sale on Y Elite?

FGC Corp
04-21-2016, 02:42 PM
I heard that Big Y is going to soon be on sale for about $460 to celebrate DNA day.
Will there also be a sale on Y Elite?

You can email us at sales [at] fullgenomes.com.

TigerMW
04-21-2016, 02:56 PM
Not for very prestigious customers I work with Mike. FTDNA is inadequate for their purposes. They can't achieve their goals at all with that limited data set.

No, not a fait accompli. Bunk. In fact, our sales have increased dramatically for a few research projects recently, precisely because of the advantages already enumerated previously by myself and by other posters in this thread.I agree that FGC has done a great job and have said so, I think even in this thread. I think FGC is particularly good for pure research. I know I am NOT very prestigious but some of your very prestigious customers get Big Y too. There is a reason for this.

What are your total number of FGC Elite tests completed/delivered and paid for in R1b by individual testers?

Kinman, Big Y is on sale right now, starting April 21st for $460.

FGC Corp, FGC should match pricing. Why not? You have the same high enrichment cost FTDNA has for Next Generation Sequencing but don't have the large I/T project and database proccessing costs nor the overhead of a physical laboratory and storage. Correct me if I'm wrong.

When you are number two you have to try harder.

FGC Corp
04-21-2016, 03:19 PM
FGC Corp, FGC should match pricing. Why not? You have the same high enrichment cost FTDNA has for Next Generaion Sequencing but don't have the large I/T project and database proccessing costs nor the overhead of a physical laboratory and storage. Correct me if I'm wrong.

Pricing is a function of enrichment costs and the size of the data capture. If one uses a smaller capture, the costs go down. Plus, if one orders a very large number of captures, then the price goes down. If you combine both options, you adopt the competitor's strategy.





I know I am NOT very prestigious but some of your very prestigious customers get Big Y too.

I'm referring to institutional projects rather than individual customers here.


What are your total number of FGC Elite tests completed/delivered and paid for in R1b by individual testers?

It is easier for me to make the research quality argument to large research customers rather than the individual customer, and that's the reality. For others, people buy based on price which is their choice. I'm not getting into the exact specifics of our sales in each individual haplogroup, but my point here is that for us making the technical case for why our product is different is a bit obscure, when dealing with an entry level customer. I understand that.

It is easier to make the case to researchers. I have done presentations for a couple research teams already myself.

TigerMW
04-21-2016, 03:29 PM
....
I'm referring to institutional projects rather than individual customers here.
...
That's excellent. We clearly need more research and you are no doubt a leader for the Y chromosome, if not the leader.

FGC Corp
04-21-2016, 03:38 PM
That's excellent. We clearly need more research and you are no doubt a leader for the Y chromosome, if not the leader.

Y chromosome sequencing is analogous to any other mass manufacturing problem. If you have the right machines and the right throughput you can beat the prices in the market. But that assumes that you can justify the short-term investment in the technical capital of machinery when that machinery will quickly become obsolete.

If you buy a $10 million xTen you need to be able to demonstrate that you'll sell 18,000 samples a year, for example.

MitchellSince1893
04-21-2016, 03:44 PM
Just my two cents from an entry level customer. I wouldn't be surprised if quite a few bigY customers end up doing Y Elite afterwards. Many of them like me will have a nagging feeling of what was missed by bigY that Y Elite would have picked up. I will know the answer to that question in about a month.

FGC Corp
04-21-2016, 03:51 PM
Y chromosome sequencing is analogous to any other mass manufacturing problem. If you have the right machines and the right throughput you can beat the prices in the market. But that assumes that you can justify the short-term investment in the technical capital of machinery when that machinery will quickly become obsolete.

If you buy a $10 million xTen you need to be able to demonstrate that you'll sell 18,000 samples a year, for example.

A counterexample would be doing something unique. If you're doing something unique, then you're not subject to the economy of scale advantages of your competitor(s). US companies largely operate on economy of scale advantages, ie Amazon, Walmart, various others, etc.

The funny thing about the economics of competition is that even very advanced services become commodity goods. Combine a unique product with a mass market appeal and one bypasses the economy of scale problem.

FGC Corp
04-21-2016, 04:12 PM
A counterexample would be doing something unique. If you're doing something unique, then you're not subject to the economy of scale advantages of your competitor(s).

In the professional world, one has the same problem. If one gets a credential widely shared,then one is easily replaceable by another with a similar credential, unless one has a unique contribution to make. Finding products and skills that are less fungible provides more security and greater profitability.

TigerMW
04-21-2016, 04:16 PM
A counterexample would be doing something unique. If you're doing something unique, then you're not subject to the economy of scale advantages of your competitor(s). US companies largely operate on economy of scale advantages, ie Amazon, Walmart, various others, etc.

The funny thing about the economics of competition is that even very advanced services become commodity goods. Combine a unique product with a mass market appeal and one bypasses the economy of scale problem.
I hope some of our voting citizens follow along with you in understanding micro-economics. Businesses who can access capital and invest in their own equipment and facilities can reduce fixed costs over the number of units sold.

The 22 immutable laws of marketing applies and you are grasping that, too. Law #1 is the law of leadership where the #1 in market share has great leverage. Law #10 is the law of division whereas if you are not #1, divide the market into niches and become #1 in market share in the new segment of the market. If you choose right, this may become the more important market over time.

All of this is to the technician's chagrin as it is better to do the right thing (be in right market in #1 place) than to do things right (execute flawlessly or in a technical superior way.) :( I'm old (obsoleted) technician.

I'll try to lay off the micro-economics as I think we've strayed enough.

EDIT: I couldn't resist continuing off tangent so I'll just edit this post to add my postscript on a sport I love, football. A great example of a technician is Peyton Manning, the Super Bowl winning quarterback. He says he was taught to "make the right decision" and then "make his decisions right". He wanted another Super Bowl and he got one, but it wasn't because he was a better player. He may not have even been functioning above average as a quarterback, at least has not been a top tier this year. He was benched. However, he made the right decision to be on a great team, the Broncos, so he got final Super Bowl victory and can ride into the sunset.:angel:

FGC Corp
04-21-2016, 04:23 PM
The 22 immutable laws of marketing applies and you are grasping that, too. Law #1 is the law of leadership where the #1 in market share has great leverage. Law #10 is the law of division whereas if you are not #1, divide the market into niches and become #1 in market share in the new segment of the market. If you choose right, this may become the more important market over time.

Our whole genome offering doesn't have a peer competitor, since there isn't another DTC offering for ancestry on the market in that area. That's an example of a different niche. That's an example of that strategy at work.

FGC Corp
04-21-2016, 04:46 PM
All of this is to the technician's chagrin as it is better to do the right thing (be in right market in #1 place) than to do things right (execute flawlessly or in a technical superior way.)

Sure. It is about making an economically broadly appealing product, or enter a market with a product that doesn't have a peer or that competitors are unwilling to enter for the time being. If one just wants to be technically correct, than one can be an academic and be exempt from economic demands.

leonardo
04-21-2016, 04:56 PM
Just my two cents from an entry level customer. I wouldn't be surprised if quite a few bigY customers end up doing Y Elite afterwards. Many of them like me will have a nagging feeling of what was missed by bigY that Y Elite would have picked up. I will know the answer to that question in about a month.

This is an appealing idea and the thought has crossed my mind. The problem is the cost. I was among the first to order the BigY and got it at their groundbreaking price (can't remember how much). After having ordered a number of ydna markers tests and autosomal tests from the big 3, my wife still wasn't very happy with the purchase (I told her it was my Christmss gift at the time). I was fortunate on two fronts: I was in the original group who sent their BAM file to YFull, so my results were analyzed at no cost, and another man who shared a sub-branch with me for the longest time had already tested with FGC and had submitted his results to YFull. So, right now, I feel as if my DNA is as covered as possible.

RobertCasey
04-21-2016, 05:07 PM
Our whole genome offering doesn't have a peer competitor, since there isn't another DTC offering for ancestry on the market in that area. That's an example of a different niche. That's an example of that strategy at work.

Another niche that is not being addressed - an economical way to test the 1,000s of private YSNPs that NGS tests generate. YSEQ offers individual YSNPs which requires too much analysis and is hard to convince others to test. FTDNA seems to be getting out of this market as they refuse to add newly discovered branches of L226. The YSEQ panels are the leading edge test since they are upgraded more often but FTDNA is making a ton of money on their panel tests.

What we need is a chip based test like CROMO2 or NatGeo that includes only private YSNPs. Earlier in your YElite rollout, you stated that you would roll out an economical way to test all the YSNPs being discovered. Since this product has significant development costs, this would require a pretty good quantity of orders that are much higher than your NGS tests. Also, you could probably only update this kind of test once a year since there are significant development costs. I think this test would sell extremely well (much better than all the individual, panel and pack tests combined). You seem to heading out towards servicing the institutional market which helps keep sales up but is changing your strategy from your original core business of the consumer market.

FGC Corp
04-21-2016, 05:59 PM
Another niche that is not being addressed - an economical way to test the 1,000s of private YSNPs that NGS tests generate.

A 4x whole genome would cover a lot of those and would also afford the upgrade option which a chip does not.

Kwheaton
04-21-2016, 06:59 PM
A 4x whole genome would cover a lot of those and would also afford the upgrade option which a chip does not.

I know you have published results for 4X for overall SNPS found but do you have a rough estimate of number of KNOWN YSNPS found?
I really would like to see some data as it might prove cost effective for project purposes.

FGC Corp
04-21-2016, 07:59 PM
I know you have published results for 4X for overall SNPS found but do you have a rough estimate of number of KNOWN YSNPS found?
I really would like to see some data as it might prove cost effective for project purposes.

As an example, 43,467 SNPs out of 54,089 SNPs are called in an example 4x kit. (Y chromosome SNPs).(using our gtype file as a benchmark).

[corrected number]

FGC Corp
04-21-2016, 08:45 PM
A 10x example (YFull assessment):

10x whole genome:
SNPs (all): 80,069
Positive: 2,762 (3.45%)
Negative: 73,354 (91.61%)
Ambiguous: 250 (0.31%)
No call: 3,623 (4.52%)

dkm1987
04-22-2016, 02:02 PM
But then you still have what Mike has mentioned previously, the relative comparison aspect. Identifying the SNPs is only one part. Case in point.
My Elite 2.0 showed many private SNPs. Not long afterward a Big Y tester came back positive for some of them as well. Several have now been vetted at YSEQ and even one is now offered at FTDNA as a single SNP. Which we have tested 9 other testers at FTDNA and the results came back yesterday proving a branch out of S1051, FGC17938.

So now we need to start testing via YSEQ some of the lower (more recent) SNPs to show and prove their status on a case by case basis. Whereas if other testers would test Big Y these comparisons would be completed without the added testing and of course costs. But of course this is the rub, the cost of either the Big Y or NGS from FGC is prohibitive so what appears to happen most often is novel SNPs are verified via YSEQ, then either individual SNPs are ordered or packs are made to help defray the costs. Which unless done via FTDNA another step still needs to be completed before the terminals are shown in FTDNA much larger database.

So the discovery is of course very important but so is the ability to easily and readily see comparisons. And even though I truly think FGC is a great product FTDNA definitely has the edge on the later aspect.

FGC Corp
04-22-2016, 02:42 PM
But then you still have what Mike has mentioned previously, the relative comparison aspect. Identifying the SNPs is only one part....

So now we need to start testing via YSEQ some of the lower (more recent) SNPs to show and prove their status on a case by case basis. Whereas if other testers would test Big Y these comparisons would be completed without the added testing and of course costs. But of course this is the rub, the cost of either the Big Y or NGS from FGC is prohibitive so what appears to happen most often is novel SNPs are verified via YSEQ, then either individual SNPs are ordered or packs are made to help defray the costs. .

Although price is always an issue, an FGC test taken in 2013 for example, remains state of the art in 2016, and FGC tests offered today are more advanced than those offered then. Our accommodation in terms of pricing consists of payment plans.

More news to come in the next months however....

Our analyses have also stood the test of time and I think there's an opportunity for substantial improvement there. The white paper that Greg presented in 2014 on our test remains true today, and the quality has improved since then.

dkm1987
04-22-2016, 09:20 PM
Although price is always an issue, an FGC test taken in 2013 for example, remains state of the art in 2016, and FGC tests offered today are more advanced than those offered then. Oh no doubt about that and again using a direct one to one comparison there were many SNPs I had in FGC Elite 2.0 that the Big Y test didn't. I do feel FGC far surpasses FTDNA in both the testing and what you get from it via the analysis but I do agree with Mike that the matching portion of why people test is integral to the entire picture, finding matches and FTDNA is truly the leader in that aspect. Then again and as you have stated, you have carved out a nice little niche and are doing that very well.

So perhaps soon the prices will come down enough that it is possible for one to economically test through both and get the best of both worlds.


Our accommodation in terms of pricing consists of payment plans.For DTC?

FGC Corp
04-22-2016, 09:29 PM
Oh no doubt about that and again using a direct one to one comparison there were many SNPs I had in FGC Elite 2.0 that the Big Y test didn't. I do feel FGC far surpasses FTDNA in both the testing and what you get from it via the analysis but I do agree with Mike that the matching portion of why people test is integral to the entire picture, finding matches and FTDNA is truly the leader in that aspect. Then again and as you have stated, you have carved out a nice little niche and are doing that very well.

So perhaps soon the prices will come down enough that it is possible for one to economically test through both and get the best of both worlds.

For DTC?


For DTC?

Yes, that's correct. Payment plans are possible for individual customers.

RobertCasey
04-23-2016, 12:25 AM
A 4x whole genome would cover a lot of those and would also afford the upgrade option which a chip does not.

4X is still $395 is not currently an economical approach. Chip array tests could be in the $100 range. This would compete directly with the existing YSEQ YSNP panel tests and FTDNA YSNP packs. The chip array could be broken up across the haplotree as well to get costs down. I think the CROMO2 would be a good approach (ten tests across the haplotree for $75 each). Since set up costs are significant, updates would be yearly at best.

ChrisR
04-23-2016, 06:59 AM
A 10x example (YFull assessment):

10x whole genome:
SNPs (all): 80,069
Positive: 2,762 (3.45%)
Negative: 73,354 (91.61%)
Ambiguous: 250 (0.31%)
No call: 3,623 (4.52%)
I would be interested in the following YFull stats of this sample:
Mean depth coverage
Median depth coverage
Length coverage
ChrY BAM file size
STRs (all) / Reliable alleles

or you may post all other YFull stats - Raw data (Y) - STRs - Novel SNPs - Raw data (Mt)

Francisco
04-23-2016, 08:15 AM
I think it is fait accompli for the average genetic genealogy tester. The Big Y train has left the station.
The Big Y advantages I summarized still apply.

1. Lower absolute entry price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

Now is a great time to jump on board as the FTDNA has their Big Y priced at $460 for DNA Day sales (April 21-26). Unless FTDNA has some magic, according to reports from FGC and TK the high enrichment costs of NGS testing mean FTDNA must hardly be making money on this. BTW, I'm a bit skeptical on how much the enrichment costs are but if they are that much, this is really a good deal.

The momentum of the train is what's important, though. Big Y orders in my projects trickle in nearly every week without special incentives. They just happen. This means critical mass in the market has been reached. These promotion events like the $460 just add an additional carload to the train.

Since genetic genealogy is only about comparisons of test results between people, the growth in the Big Y database is critical. We will find people and surprises we didn't know about.

I think FGC Elite looks like a great test, but even if you do it (and assuming you have that kind of money) you should consider getting Big Y anyway to get registered in the Big Y database and find people as they find you. Many people have done both tests. It may seem crazy but if you are avid and can afford it, it's worth it. These folks are just true pioneers.

A fellow poster John estimated the Big Y penetration based on some major R1b haplogroup project estimations I've done. His numbers are rough but he had L21 at 40% Big Y penetration of the total L21+ confirmed people [EDIT: I recalculated and think it should be 30% for L21]. He had L238 at 60%. He had U106 at 25%. I've got DF19 at 69%. I think DF19 is the winner with 145 Big Y's out of 210 people.

The net is there getting up to several thousand people in R1b alone with Big Y results in their database.

Hey, not so easy not so cheap.
FTDNA donīt sale directly the Big Y to "new customers", we must buy some other test first, letīs say a 129u$s Y37STR good for nothing (as Big Y has really more STRs than this test, but FTDNA donīt want their customers to get them "for free" first).
So, we should first pay more (589u$s) and then have to wait to 2 tests to be done; and that NOT counting that Big Y donīt give also (as they donīt give the 300+STR FGC gives -we must pay a third party analysis service to read the .bam to get them!) the mithocondrial info that charges today "at sale" for 149u$s
So with the "big offer" we must pay 738u$s for the same product that FGC sells (Y SNPs, STR, mDNA) but with much less quality (in SNPs). And without this special offer, the sum gets at least 100u$s more (and that is more than FGC charges!!).
The only real vantage for FTDNA is to have a captive market.

They will be a "fait accompli" perhaps when their could offer: Big Y + STRs + mDNA at those 450/500u$s (and not for the top tier who wants the best of the best in the market: Whole genome, or Full Y at least)

FGC Corp
04-23-2016, 12:55 PM
I would be interested in the following YFull stats of this sample:
Mean depth coverage
Median depth coverage
Length coverage
ChrY BAM file size
STRs (all) / Reliable alleles

or you may post all other YFull stats - Raw data (Y) - STRs - Novel SNPs - Raw data (Mt)


Here:
10x whole genome
0.18 Gb .BAM file size
87.54% Length coverage
5X Median depth coverage
7999X MAX
11.4X MEAN
1X MIN

30x whole genome:
ChrY BAM file size:
0.29 Gb
Reads (all):3,810,713
Mapped reads:3,808,394 (99.94%)
Unmapped reads: 2319 (0.06%)
Length coverage: 22,858,252 bp (89.10%)
Min depth coverage:1X
Max depth coverage:7999X
Mean depth coverage:21.38X
Median depth coverage:12X
No call: 2,795,314 bp

Average Callable Loci (Y chromosome)

Whole genome samples, build 37 reference:
30x 14,558,001
20x 13,888,138
15x 13,745,646
10x 8,046,540
4x 1,050,996
2x 349,397

Y Elite 2.0, Batch 9007, build 37 reference:
14,530,352

Y Elite 1.0, build 37 reference
14,073,254

FGC Corp
04-23-2016, 04:57 PM
Hey, not so easy not so cheap.
FTDNA donīt sale directly the Big Y to "new customers", we must buy some other test first, letīs say a 129u$s Y37STR good for nothing (as Big Y has really more STRs than this test, but FTDNA donīt want their customers to get them "for free" first).
So, we should first pay more (589u$s) and then have to wait to 2 tests to be done; and that NOT counting that Big Y donīt give also (as they donīt give the 300+STR FGC gives -we must pay a third party analysis service to read the .bam to get them!) the mithocondrial info that charges today "at sale" for 149u$s

Somehow this fact is consistently omitted in a variety of price comparisons. I wonder.

Petr
04-23-2016, 10:31 PM
Hey, not so easy not so cheap.
FTDNA donīt sale directly the Big Y to "new customers", we must buy some other test first, letīs say a 129u$s Y37STR good for nothing (as Big Y has really more STRs than this test, but FTDNA donīt want their customers to get them "for free" first).
So, we should first pay more (589u$s) and then have to wait to 2 tests to be done; and that NOT counting that Big Y donīt give also (as they donīt give the 300+STR FGC gives -we must pay a third party analysis service to read the .bam to get them!) the mithocondrial info that charges today "at sale" for 149u$s
So with the "big offer" we must pay 738u$s for the same product that FGC sells (Y SNPs, STR, mDNA) but with much less quality (in SNPs). And without this special offer, the sum gets at least 100u$s more (and that is more than FGC charges!!).
The only real vantage for FTDNA is to have a captive market.

They will be a "fait accompli" perhaps when their could offer: Big Y + STRs + mDNA at those 450/500u$s (and not for the top tier who wants the best of the best in the market: Whole genome, or Full Y at least)
It's not so straightforward.
1. I was able to order just BigY, without STR, just asked the customer service.
2. The cheapest STR test is Y-DNA12 at $59: https://www.familytreedna.com/group-join.aspx?Group=Project_Pending
3. My Y Elite 1.0 analyzed by YFull shows 32/37; 57/67; 97/111 STRs, my BigY analyzed by YFull shows 33/37; 55/67; 97/111 STRs, Y Elite 1.0 analyzed by FGC shows 27/37; 50/67; 75/111 STR markers.
4. The coverage of mtDNA in Y Elite 1.0 is not perfect, 19 bp no call, 26 bp one reading.
The result for Y Elite 2.5 may be different.
Y Elite is much better than BigY for sure.

FGC Corp
04-24-2016, 12:40 AM
3. My Y Elite 1.0 analyzed by YFull shows 32/37; 57/67; 97/111 STRs, my BigY analyzed by YFull shows 33/37; 55/67; 97/111 STRs, Y Elite 1.0 analyzed by FGC shows 27/37; 50/67; 75/111 STR markers.


Y Elite 2.0:
107 /111 STRs per YFull.

FGC Corp
04-24-2016, 03:52 PM
4X is still $395 is not currently an economical approach. Chip array tests could be in the $100 range. This would compete directly with the existing YSEQ YSNP panel tests and FTDNA YSNP packs. The chip array could be broken up across the haplotree as well to get costs down. I think the CROMO2 would be a good approach (ten tests across the haplotree for $75 each). Since set up costs are significant, updates would be yearly at best.

The problem with the design of a custom chip is that while it would be possible to design a new chip with all of the current ISOGG SNPs on it, as a practical matter that would require a guaranteed volume to ensure that that $100-$200 price range could be achieved. Yes, in principle, that could be done. One could add 100,000+ SNPs to an existing design (or more).

Francisco
04-24-2016, 04:03 PM
It's not so straightforward.
1. I was able to order just BigY, without STR, just asked the customer service.
2. The cheapest STR test is Y-DNA12 at $59: https://www.familytreedna.com/group-join.aspx?Group=Project_Pending
3. My Y Elite 1.0 analyzed by YFull shows 32/37; 57/67; 97/111 STRs, my BigY analyzed by YFull shows 33/37; 55/67; 97/111 STRs, Y Elite 1.0 analyzed by FGC shows 27/37; 50/67; 75/111 STR markers.
4. The coverage of mtDNA in Y Elite 1.0 is not perfect, 19 bp no call, 26 bp one reading.
The result for Y Elite 2.5 may be different.
Y Elite is much better than BigY for sure.

1. That`s true, but they donīt show that. Why? They expect NEW customers to pay more tests. And of course, a new customer donīt know, cause case he is new to FTDNA and their tests offers. They see only what FTDNA shows in their web
2. The 12 STR is not offered in any place of the FTDNA web, the user must navigate to the groups to then see that. A new customer will do that? of course not, he will just start buying the cheapest in offer (STR37 at 149$ in "DNA day offer").
You can check the page and their offers:
https://www.familytreedna.com/products.aspx
No Big Y, no 12STR at 59$... you go to learn more about the Y tests?

https://www.familytreedna.com/y-dna-compare.aspx
and... no 12STR at 59$ and no Big Y there.
As you said to put those products in the cart, first you must call the "customer service". And that is preposterous and made to grab money for the new users, as they will never call "customer service" for a product they do NOT know exists and is not shown in their web site.

4. mtDNA is not perfect, sure as http://www.it2kane.org/2015/11/ngs-y-chromosome-alignment-on-grch38/ shows it will only gives a coverage of 95/96%. I guess thatīs a very good number for most people. And Big Y gives zero (they hide this data, that has a not so good coverage of only 72%). But, well if you says itīs not enough, we must admit that this mtDNA at least give more value than the "mtDNA +" from FTDNA at 69u$s.
So STR12 + mtDNA+ is 128u$s that we must add to their Big Y to make a fair comparision with the FGC

Finally, one thing with FGC is that each year they get a better test and are moving forwards (from GRCh37 chromosome alignment to GRCh38, new enrichment process, from 100bp to 250bp, new quality controls). And there it is, FTDNA sleeping and not updating their Big Y for a long time, hey! they even get this test to the door only when they saw FGC as a threat in the market... otherwise, people would only do the "money grab ladder" (going from 37STR to 67 to 111, adding mtDNA, "walking the Y" and purchasing the SNPs one by one or in packs, paying A LOT).

The only reason I donīt get this FGC test yet is problems with the credit card (that must allow a buy in a foreign country and foreign exchange, with a sum that is BIG for my country). The good, is that all my waiting brings more technologies and know how to their tests (and they reduce their prices too).

*And Y Elite 2.5? Thatīs new for me, I only hear about the Y Elite 2.1b version as their last upgrade. What is new in the 2.5 flavor?

FGC Corp
04-24-2016, 08:44 PM
I think it is fait accompli for the average genetic genealogy tester. The Big Y train has left the station.

Ultimately, I don't take the train anymore. I prefer jets.

Amerijoe
04-24-2016, 10:38 PM
Ultimately, I don't take the train anymore. I prefer jets.

TOUCHÉ

MitchellSince1893
04-25-2016, 12:39 AM
I had forgotten this, but back when I ordered BigY, FGC was an unknown quanity...at least to me. I just wasn't ready to risk that sum of money on an "upstart". I wondered how long it would be around.

That was almost 3 years ago and FGC is still going strong...improving and expanding their product line. I think FGC has proven itself as providing quality products for the money. Sure the wait can be long but IMO it's worth it.

There's nothing wrong with going the BigY route, but I don't think you will regret going with FGC. Either way you will learn more than you will with single SNP tests or chip based ones.

kinman
04-26-2016, 01:00 PM
A new updated (version 4.4) of the YFull tree has appeared: https://www.yfull.com/tree/

TigerMW
04-26-2016, 02:19 PM
The numbers don't lie. What I posted below is happening and has happened. Critical mass has been achieved. There are about 200 new Big Y orders in the R1b projects I can see in the last four and a half days. This is only the projects I can see and only R1b. There are now thousands of R1b Big Ys in the database, with the results integrated and accessible to projects and matching systems. These are not just anonymous academic results but real people who can be contacted for genetic genealogy purposes, including recruiting to take other tests, be they at FGC or YSEQ or whatever.

Today is the last day of the sale price $460. Keep in mind that competitors have said enrichment costs per test can't run much less than $400 so this is a good deal.

"I think it is fait accompli for the average genetic genealogy tester. The Big Y train has left the station.
The Big Y advantages I summarized still apply.

1. Lower absolute entry price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

Now is a great time to jump on board as the FTDNA has their Big Y priced at $460 for DNA Day sales (April 21-26)."

FGC Corp
04-26-2016, 02:23 PM
"I think it is fait accompli for the average genetic genealogy tester. The Big Y train has left the station.
The Big Y advantages I summarized still apply.

No it hasn't Mike. In fact, our business has improved on our side this year.

TigerMW
04-26-2016, 03:26 PM
No it hasn't Mike. In fact, our business has improved on our side this year.
That's very good and I wish you well. I'm not saying FGC is on a death knell. I'm just saying the Big Y is the clear market share leader and its growth has legs of its own. That the momentum of the train. That's fait accompli. The advantage in size of the matching database of real people and the corresponding genetic genealogy systems (projects, surnames, etc.) is of real benefit to people. It is not something easily overcome.

As I said, this is to the chagrin of the technician.

I agree with you that it makes sense for you to divide the market with a new niche and claim leadership in that niche.

leonardo
04-26-2016, 03:33 PM
I can see the appeal of both. As I posted earlier, I had the best of both worlds, so to speak, with a match having tested at FGC, then submitting their results to YFull. So, when I received my BigY results and submitted them to Y Full, I was Able to be part of a new branch on the tree. Just yesterday, another match who had taken the BigY test and submitted the results to YFull resulted in a TMRCA of 700 years. Not bad for ydna. By the way, YFull continues to increase the results it offers. It is well worth the $49, regardless of which test one takes. Its data base is becoming impressive as well.

RobertCasey
04-26-2016, 03:34 PM
1. Lower absolute entry price
2. Faster test processing time
3. Convenience
4. Size and access to the matching database (using the same testing method/coverage)
5. Early feeder of your SNPs into a broader set of offerings
6. Better prospects as a proven on-going, self-funding concern

1. Yes, the Big Y is around 30 % less than Y-Elite but covers 30 % less of the YCHR (I still have 12 private YSNPs from Y-Elite 1.0 not tested by Big Y). With your rationalization, you should stop recommending testing of higher resolution 111 marker YSTR tests based on the fact that higher coverage has no value ?? Also, if costs are that important, why advocate no testing of private YSNPs. Under L226, testing of private YSNPs yields new major branches at 10 % of the cost of Big Ys to date. So you advocate a strategy that of all Big Ys (a 30 % savings for 30 % less resolution) but encourage a strategy of only Big Ys which currently costs 10X when compared to private YSNP testing. There needs to be mixture of high resolution testing (at least 10 %) and 75 % of the branches can be found by individual YSNP testing.
2. Faster processing - I grant you that - but FGC is continually improving their product (coverage and quality) and they are smaller company. This is a serious disadvantage that FGC needs to close the gap on. But sometimes the wait for higher resolution tests are worth it (ie. FGC high resolution tests and the eventual 111 marker upgrades) ??
3. Convenience - This is another plus for FTDNA. It is much easier to get individuals to test with FTDNA since their role in YDNA is well established. Also, FTDNA offers SNP packs which complement their offering as well as their YSTR database which is key for research. However, you now have to go to YSEQ for individual YSNP testing of private YSNPs which is a major negative trend. Another convenience factor which probably driving this is the data extraction piece. Combining data from multiple sources is not convenient and diverts time from other useful activities. But usually convenience of ordering all from one source eventually leads to less innovation and higher costs. Notice the trend - FGC NGS tests (FTDNA drops Walk the Y and comes back with initial Big Y results around a year later). YSEQ SNP panel tests - now we have robust SNP packs from FTDNA. Testing of private YSNPs is too low quantity, so that product line is winding down because testing individual private YSNPs is not worth their time (increasing our costs by 10X for the NGS tests only strategy).
4. Databases - The size of their YSTR database is extremely important and their YSNP reports are an important resource as well. Also, their internal database of Big Ys have also allowed FTDNA to reclaim their leadership of haplotree development (but they were bad here for many years and I think Alex's Big Tree and YFULL helped improve their hapotrees). But for NGS tests, most leading edge researchers use BAM files which reveal much more and are much more accurate than FTDNA generated files. So their database plays very little role in analyzing the NGS BAM files. These are done manually by most of the haplogroup researchers (hopefully Alex will be able to continue and expand his work which is much more accurate than the FTDNA haplotrees).
5. Broader range of offerings. Again, this is another major convenience factor. But you still have to go to FGC for higher resolution NGS tests, you still have to go to YSEQ for more current pack/panel tests and now your only option for individual YSNP testing is going with YSEQ since FTDNA is now longer adding any private YSNPs.
6. This refers to convenience again. Along with the convenience of buying from one source - eventually yields a slower pace of innovation (higher costs) and a lack of competition is not great (higher costs).

Again, I think FTDNA plays the key role in being the leader. But I am getting pretty tired of having to go other vendors for: 1) YSEQ for individual YSNP testing of private YSNPs that discover new major branches at 10 % of NGS tests. These tests are being dropped by FTDNA for lack of market place (and loss of profit from Big Y); 2) having to go to YSEQ for the most current YSNP panel/pack test - FTDNA always lags in this market since their tests are based on equipment that produces output at lower costs but require more setup (profits over innovation of being kept up to date). 3) having to go to FGC for higher resolution testing for YSNP discovery who continues to innovate by improving their products whereas FTDNA's product is static. Walk the Y was still being offered way too long and a small upstart can be them to the market by a year ???

leonardo
04-26-2016, 03:46 PM
3. Convenience - ...

I see one of the major conviences of FTDNA as not having to offer up another DNA sample. I have had a difficult time getting relatives to swab or spit. If one offered up a sample for an autosomal or ydna marker test, they can then purchase the BigY without any more effort.

RobertCasey
04-26-2016, 04:59 PM
I see one of the major conveniences of FTDNA as not having to offer up another DNA sample. I have had a difficult time getting relatives to swab or spit. If one offered up a sample for an autosomal or ydna marker test, they can then purchase the BigY without any more effort.
Remember, I am only recommending 10 % test at higher resolution, so if getting another DNA sample (or the myriad of other valid exceptions) still goes into the 90 % route.

My frustration under L226 is that my original FGC Elite 1.0 test (first NGS test under L226) has produced the three major branches in the trunk of the L226 haplotree. The trunk of the tree is FGC5660>FGC5628>FGC5659 since 30 % of the private YSNPs are stuck at the trunk with FGC5660*, FGC5628* and FGC5659*. Even though we now have 24 known branches, 30 % of the private YSNPs are stuck in these three branches. After 48 Big Y tests later, no new branches in the trunk of the tree. If the rate of discovery remains the same ratios that Big Y tests have revealed, my 12 private YSNPs that are untested by Big Y should find one more major branch and 3 or 4 more private genealogical YSNPs as well. With even higher resolution Y Elite 2.1, another major branch in the trunk may be possible. After 48 Big Y tests, it is very unlikely that testing more lower resolution tests (similar to 67 markers tests) will yield more branches in the trunk.

JamesKane
04-26-2016, 05:58 PM
We've seen Big Y peform well until the branches hit 1200 to 1500 AD. The test however lacks sufficient coverage resolution to attack L270.2 or A542. Both resist branching more recently despite having more than 5 tests which should show something. Both would be better served with a different platform than investing in more Big Ys.

The train is ultimately conducted by project admins and lecturers. It is miscommunication coming from us that funnels everyone to the same platform. Statements of Big Y being the only option to discover your paternal line are patently false. FGC has multiple products that do this as well. The new Veritas Genetics WGS test will do this. The important thing is people need education that there are options. It is up to them to make the value judgement.

As to FTDNA's database... Just how useful is it really when most need to send their results off to 3rd parties to make sense of the results. The reporting and databases maintained by project analysts is an indictment against that argument. When Alex and to some extent myself are no longer being sent the files as first reaction when a test is returned FTDNAs SNP data base size might have merit. The STR database really isn't relevant to discussions of NGS.

IanFitzpatrick
04-26-2016, 06:28 PM
James,

As you well know, if it was not for the incredible work of a group of people outside of FTDNA, yourself and Alex included, I would still be back at FTDNA's listing of L21 terminal HG and would have paid good money for very little return. I do not understand why this is so?

And, since I am listed as just L21, the list of FTDNA matches I have is currently at 1300 and growing daily, only 2 of these matches are relevant which really makes the whole use of processing power used to compile the database questionable at best.

The opportunity for real discovery in this field is being held back in my opinion by everything you stated in your "on the nose" post above. Is this just a question of the business side of DNA analysis being what it is to make money or a true lack of organizational structure within a company to provide a truly meaningful service? Or, maybe they are happy with the status quo?

My case is a prime example of this.

FTDNA = L21

Private analysis = R-P312/S116 > L21/S145 > DF13 > FGC11134 > ZZ44 > 19567651-T-C > 7716880-A-G > 8858535-G-A with one very strong surname match

I also see that many truly dedicated project managers seem somewhat afraid to "rock the FTDNA boat", which I understand completely if this is indeed the case. If it is so, that is just sad.

TigerMW
04-26-2016, 08:50 PM
I hope everyone is noting that I wish FGC Corp well and and am very glad to see them enter the market and create sorely needed competition. That does not mean I have to drag other vendors through the mud to be politically correct. All of the vendors have warts. I'm also still not criticizing others personally, but that doesn't mean I give in to political correctness.

The facts are still the facts and Big Y's market share leadership leads to value in and of itself.

I think it is fait accompli for the average genetic genealogy tester. The Big Y train has left the station.
The Big Y advantages I summarized still apply.
...
1. Lower absolute entry price
...
4. Size and access to the matching database (using the same testing method/coverage)
....

I was thinking about this as we see if FTDNA is up to anything as far as enhancements to Big Y. Last fall, Bennett Greenspan said there would be enhancements and they would apply to existing customers. I don't know what he was talking about, but as we speculate I have to say I'd rather have a lower price than more coverage.

Think about it. We see people taking more expansive coverage tests and that's good and the pure researchers and technicians really like that. However, it's like having a tsunami flood your island with data. A bunch of private SNPs don't do you much good. You have to have more people test for your private SNPs. That means you have to go find them and convince them to probably buy a custom SNP panel. This ends up being you have to convince other people to test for your SNPs. Perhaps the situation warrants that you have to fund tests of other individuals for your SNPs on them.

We just need more people testing NGS testing and since Big Y has already set the market share standard for coverage, that's what we need covered. Comparisons are most valuable between people tested for the same locations with the same methods. Hence, we need lower pricing, not more coverage. Coverage is good, please don't misunderstand me. I just think the absolute entry price of NGS testing is way too high and that is the real inhibitor.

If we have a bunch of private SNPs on our island, how are we going to find those people to play our technical sandboxes? FTDNA's STR database, which most of us participate in, is one place. There is nothing comparable unless you are talking about scraping their project screens but that doesn't really get it and is out of date constantly. Yes, for recruiting for NGS tests and custom panels "the STR database really" IS "relevant".

I like surprises, new learnings or what some would call insights. I'm a little lazy so I like people to find me as well as me find them. If a person from who knows where takes Big Y and matches me well, they will show up in my FTDNA Big Y database matching screens. I would never say Big Y matching screens are easy to use. They are counter-intuitive, but regardless you can find people (living people who paid for their own tests) and contact them to try to share information in a phylogenetic analysis, which you can later feed back to FTDNA's haplotree or to FGC/YFull interpretations, the Big Tree, etc.

Big Y test results do integrate and show up in project and matching screens.

FGC Corp
04-26-2016, 09:39 PM
We've seen Big Y peform well until the branches hit 1200 to 1500 AD. The test however lacks sufficient coverage resolution to attack L270.2 or A542. Both resist branching more recently despite having more than 5 tests which should show something. Both would be better served with a different platform than investing in more Big Ys.

Greg made the point about the differences between Big Y and Y Elite coverage in January 2014. In the interim those findings have been substantiated by a variety of people, and I believe you've duplicated our findings via your own independent work. At the time, some questioned Greg's integrity on that point and it was quite wrong.

TigerMW
04-26-2016, 09:56 PM
... Again, I think FTDNA plays the key role in being the leader. But I am getting pretty tired of having to go other vendors for: 1) YSEQ for individual YSNP testing of private YSNPs that discover new major branches at 10 % of NGS tests. These tests are being dropped by FTDNA for lack of market place (and loss of profit from Big Y);I agree that FTDNA has to re-straighten its act out on this. They committed last year (to me anyway) that SNPs on the haplotree will be available on via Advanced Tests/Sanger Sequenced but they have not followed through yet. I think they might, but they'll have to complete their first round of SNP Packs is my guess. In my own little clade I am very fortunate. I pumped as many of the early bird Big Y shared/public SNPs from my clade in their system and most have been available on Advanced Tests for a year and a half (before they realized they had a tsunami problem.)


2) having to go to YSEQ for the most current YSNP panel/pack test - FTDNA always lags in this market since their tests are based on equipment that produces output at lower costs but require more setup (profits over innovation of being kept up to date). I don't think that FTDNA will have ever be anywhere close to agile like YSEQ, although, probably across all haplogroups, FTDNA will provide support. I don't use the word "better" support but they position that they cover marketplace so this is good for all of the haplogroups.
This is one bit of hope here for the FTDNA SNP Pack style of testing. Since it is all in one fell swoop they can truly blast through phylogenetic equivalent blocks that are lightly tested. The result is that they don't need as many updates since packs are not tree traversing (dependent on the tree as we know it.)


3) having to go to FGC for higher resolution testing for YSNP discovery who continues to innovate by improving their products whereas FTDNA's product is static... I agree that FGC's products play an important role, particularly for advanced research. I agree that FTDNA needs some upgrades, but as I've stated, I think the bigger deal for Big Y would be a substantial price cut. I know the competitors say there is a floor so today's pricing is about as good as you will get but that doesn't mean that lower prices aren't needed.

IanFitzpatrick
04-26-2016, 10:02 PM
Mike,

I get what you are saying, however, without individuals outside of FTDNA the results from BigY would be about as useful and the Family Finder test when it comes to any real genealogical value.

I have 2 matches out of 1300 total that are relevant if I went by FTDNA results, privately the Big Tree has built a whole new branch. Who are the people that are pushing the genetic tree forward?

This matching mechanism and growth of the genetic tree by all appearances is not being driven by FTDNA when it really should. I do not understand how they do not see this as being a good business to be in control of this part of a growing field?

TigerMW
04-26-2016, 10:24 PM
... I get what you are saying, however, without individuals outside of FTDNA the results from BigY would be about as useful and the Family Finder test when it comes to any real genealogical value....
I absolutely agree with you that volunteer hobbyists of all sorts are key to progress. It's true that FTDNA receives benefits from volunteers and they have not supported them as well as they should. The volunteers are also useful across the board to other vendors. FTDNA's customer database is also useful to volunteers and competitive vendors, as exposed through the project management system and FTDNA GAP tool.

It is not unusual for third party niches to fulfill needs in any industry. I think the FGC and particularly the YFull interpretations and supporting tool set have been helpful. This is a good thing. We don't want FTDNA to do it all themselves. It wouldn't go well. I don't really don't want them doing phylogenetic analysis. They should develop much better tools to support it, though.

The truth is, I'd be happier if they just made their on-line tools extra slick and powerful AND lowered pricing. Lower pricing means more individuals engaged which is good for everybody, including the niche competitors. They just have to find, define and defend their niches.

IanFitzpatrick
04-26-2016, 10:48 PM
Mike,

I am looking at this through the lense of the consumer, as just a neophyte in the realm of genetic genealogy who has exhausted the paper trail and wants to go further

This is what FTDNA sells

"Family Tree DNA is dedicated to helping genealogists find lost relatives when the paper trail hits a brick wall. Our service was created for the serious amateurs and the professional genealogists who wish to extend their family trees by confirming a link where no conventional source records exist."

I felt let down and almost ripped off after taking the BigY, I paid for the quoted service and by no means feel I received this from FTDNA.

Now, I would say if I was FTDNA, it would concern me that

A: a group of "volunteers is kicking my butt at what I should be providing to my customer

B: that this group forms its own company to charge for this service and the testing goes elsewhere

Also, as I know you are involved in L21 so imagine you are from another group, the L21 group is by far the most well serviced group by the excellent volunteer community, I am sure glad I am I this group.

FTDNA needs to really look hard at what they want to provide to the consumer and be good at that or someone will come along and knock them of the pedestal real quick, it happens when there is money to be made

TigerMW
04-26-2016, 10:58 PM
...
Who are the people that are pushing the genetic tree forward?

Your favorite project admins and volunteers are pushing the tree forward and FTDNA has a Y team including a fairly new assignee who has been very responsive. I don't have time to do the checks again but several months ago I actually compared ISOGG's tree to FTDNA's haplotree for R1b. FTDNA had many, many more branches. FTDNA is not relegated to NGS tests for their haplotree so it is bound to be larger than the NGS only trees like the YFull's. This is spotty by subclade, and where FTDNA has developed SNP Packs. That is much of the reason I've pushed hard on getting FTDNA to get out complete coverage of R1b subclades via their Packs. If it goes into Pack it is automatically evaluated for their haplotree. It's not there yet, but growing in leaps in bounds.

If you have a Big Y test and match someone with a new terminal public SNP it's only a matter of time before your haplogroup label gets updated. It just happens (with the Y team and submissions made to them.)

Also, if you have Big Y and our in the Big Y database your new Fitzpatrick SNPs have a good chance of consideration for future packs and Geno2 (I'll cover my ears on including that one today). I think that's what you want. Your SNPs tested by a lot of folks.

I see you are new to the forum. There is an FTDNA section (under commercial) if you want to complain about them and their marketing. No one says that FTDNA is making a lot of money on Big Y. It appears to be the opposite if FGC and YSEQ principals are correct.

IanFitzpatrick
04-26-2016, 11:10 PM
Thanks Mike, I have made FTDNA aware of how I felt and it fell on deaf ears.

I just wanted to make a point here on how important all of the volunteers are, including yourself and I feel that FTDNA does not even respect the work you all do and they should!

JamesKane
04-26-2016, 11:34 PM
Hence, we need lower pricing, not more coverage. Coverage is good, please don't misunderstand me. I just think the absolute entry price of NGS testing is way too high and that is the real inhibitor.

I'll just reply to this piece. Not every person considering testing for genetic genealogy purposes is price sensitive. They are looking for solutions which will yield them the data necessary to learn more about their ancestry and possibly some health information. While Big Y has been on the market the price delta between FGC's Y test has totally changed. As I noted before unless Big Y is on sale, Y Elite has better overall bang for the buck. We've also seen new WGS options from FGC and now Veritas. These tests are the future of genetic genealogy not the dedicated Y tests.

Yes, a lot of folks are price conscious. A lower cost lower coverage Y test might be the answer or perhaps the 2x or 4x WGS tests. I don't see the current tests on the current platforms getting less expensive.

FGC Corp
04-26-2016, 11:34 PM
Now, I would say if I was FTDNA, it would concern me that

A: a group of "volunteers is kicking my butt at what I should be providing to my customer

B: that this group forms its own company to charge for this service and the testing goes elsewhere



I just noticed that comment. FGC consists of former FTDNA project managers and customers who wanted Y chromosome and whole genome NGS testing made available. That's us.

jbarry6899
04-26-2016, 11:42 PM
I feel that FTDNA does not even respect the work you all do and they should!

I have been a project administrator for some time, and I must say that I have never felt that FTDNA did not respect our work. I haven't always agreed with their business decisions, and have sometimes been frustrated by delays or lack of features that might have made our task easier, but I have found them to be generally responsive and helpful. I am currently working closely with one of their lead technicians on a unique ancient DNA project that ties into our surname research. She has been diligent, dedicated, helpful and extremely knowledgeable. I also have had a number of occasions to ask for assistance from Jeannine and Jim, the group liaisons, and they have taken my concerns seriously and worked toward solutions.

I've also had very good experiences with FGC, but for our purposes, being able to integrate BigY results into the FTDNA project infrastructure has been essential.

Of course, as they say, your mileage may vary!

Jim

FGC Corp
04-26-2016, 11:53 PM
I'll just reply to this piece. Not every person considering testing for genetic genealogy purposes is price sensitive. They are looking for solutions which will yield them the data necessary to learn more about their ancestry and possibly some health information.

One example that we have is that a large project identified a few SNPs that they wanted called with greater reliability. We did a redesign of the test, i.e. Y Elite 2.1, that addresses that request. Calling those SNPs stretches the limits of current technology.

miiser
04-27-2016, 12:06 AM
In terms of the test itself, FGC is unquestionably superior in both quality and value. The component missing from FGC is the community and communication channels.

With FTDNA, customers are already within the testing community, usually as the result of a previous STR test and membership within a surname project and/or haplogroup project. This provides a channel for testers to communicate with each other and receive guidance from administrators. People hear of NGS through this channel, and will then distribute their NGS test data for analysis, incorporation into Alex's Big Tree, discussion within various forums of the meaning of the branching, and so on in order to hopefully get some meaningful genealogical information from their test.

FTDNA is the gateway into the community. For testers who are already within the community, FGC is a superior product.

FGC's biggest challenge is that people don't typically know anything about NGS until after they become members of the FTDNA community. For potential new customers not yet within the community, FGC lacks the infrastructure to bridge this gap. There is no advertising and no centralized location to collect all the data for analysis and discussion.

What really needs to happen is for something along the lines of Alex's Big Tree to become a formalized "company" with robust software - a mostly automated process for uploading the data, analyzing the data (both SNP and STR), and incorporating it into a tree. And this service needs to include a mandatory account setup process which includes lineage information that can be viewed by matches.

I don't really care whether this service happens within FTDNA, FGC, or some outside company. But this software analysis is the most important aspect of product differentiation. And right now, it's not being done effectively at either FTDNA or FGC. Whichever company wants to be the market leader 5 years from now needs to focus on software development. Once the software and online community is fully established, the test itself gets reduced to a humble commodity. The only thing that matters after this point will be the cost per SNP discovered. The software company with the useful database, in the position of control, will then become the middle man distributor of the NGS testing services, who will be forced by competition to provide similar products at the same price with very narrow profit margins.

FGC Corp
04-27-2016, 12:15 AM
But this software analysis is the most important aspect of product differentiation. And right now, it's not being done effectively at either FTDNA or FGC. Whichever company wants to be the market leader 5 years from now needs to focus on software development. Once the software and online community is fully established, the test itself gets reduced to a humble commodity. The only thing that matters after this point will be the cost per SNP discovered. The software company with the useful database, in the position of control, will then become the middle man distributor of the NGS testing services, who will be forced by competition to provide similar products at the same price with very narrow profit margins.

Yes, we are aware of this and do have significant upgrades to the user interface and graphic interface under development, which should be ready soon.

lgmayka
04-27-2016, 10:09 AM
We've seen Big Y peform well until the branches hit 1200 to 1500 AD.
This is a crucial point.

- Most clades across the human haplotree have not yet been extended down to 1200 A.D. For all these clades, Mikewww's suggestion is exactly correct: We need as many NGS testers as possible, which in turn requires the lowest possible test price--i.e., the Big Y on sale.

- A few clades, presumably mostly of British Isles and Ashkenazi Jewish ancestry, have already been extended down to 1200 A.D. and later. These clades need higher precision, whether through a higher-quality NGS or individual SNP tests.

JamesKane
04-27-2016, 11:24 AM
I'm not disagreeing, Igmayka. Food for thought though.

Most testers appear to be looking for a connection in the 1700 to 1800 AD range. How will those going with the absolute lowest price to fill out those older branches test the remaining 7,000,000 bases left totally out of their results when it comes time to move into modern times? I know everyone proposes that STRs will pick up the slack here, but remain skeptical. STRs are a method for making correlations on large populations. The math starts to break down dealing with individuals. Further the cost of Big Y and the 111 marker set is more than the options that give you the more complete Y sequence.

RobertCasey
04-27-2016, 01:51 PM
Most testers appear to be looking for a connection in the 1700 to 1800 AD range. How will those going with the absolute lowest price to fill out those older branches test the remaining 7,000,000 bases left totally out of their results when it comes time to move into modern times? I know everyone proposes that STRs will pick up the slack here, but remain skeptical. STRs are a method for making correlations on large populations. The math starts to break down dealing with individuals. Further the cost of Big Y and the 111 marker set is more than the options that give you the more complete Y sequence.

I only got heavily into genetic testing since I have been a hard core genealogist for forty years now. My Casey research has hit the brick wall in western South Carolina in the mid 1700s. We know there are at least 40 different males lines in this time frame where little progress has been made over the last 40 years and speculation continues to turn into compiled family histories. With only YSTRs, the progress has been pretty minor but important: 1) We have confirmed that around twenty of these lines are indeed related but did find one outlier Casey living in the same area; 2) We are very fortunate to have very isolated genetic cluster with seven mutations since L226 creation which is only 1,400 years old (my ancestors rolled the dice and got 9s to 10s for long period of time). Our unique signature under the very isolated L226 signature can be assigned to our surname of one Casey progenitor. 3) Along the way, we have picked up two interesting NPEs to add to our analysis (one NPE is supported by probate documentation). 4) we discovered one major branch of a very rare marker value that divides the entire cluster into two 50/50 branches (this is pretty rare and lucky). 5) We have discovered one fairly close match that partially matches our signature but is obviously more distantly related to our cluster (its surname is Kersey which is a close match to Casey). The Kersey tester has a pending test of four private YSNPs that tested positive by others in our cluster.

But testing YSTRs alone is not making as much progress as anticipated by this community of researchers which have been exchanging traditional research for several decades and a lot of the researchers/sponsors are aging. Enter FGS tests. We now have our YSNP testing down to our surname cluster with the discovery of the new L226 branch FGC5639 which has only tested positive within our cluster of testers (after 48 Big Y tests and my one FGC test). Only one $39 individual test discovered this branch where 48 NGS tests failed to discover this branch. But it has been frustrating to get the L226 community to take any FGC tests which has two major advantages: 1) one for genealogical clusters, only FGC tests can produce the maximum number of private YSNPs to break down brick walls after surname creation (around 1,000 years ago for my Irish clan). 2) There are major bottlenecks where large numbers of private YSNPs are stuck under major branches. With the 24 L226 branches discovered to date, around one-third of all private YSNPs are attached to only three branches of the 24 known branches. Again, only higher resolution tests will break up these bottlenecks since 48 subsequent Big Y tests have failed to discover even one new branch in the trunk of the L226 haplotree that contains all three bottlenecks.

I convinced the most remote Casey line from my FGC Casey tested line to take the FGC test to discover more private YSNPs. He will be the only NGS tester to have show results on 12 my additional private YSNPs that 48 Big Y tests have not tested. Since his test is Y Elite 2.1 vs. my test which was Y Elite 1.0, he will hopefully find two or three additional private YSNPs with more coverage. Also, since we share no common ancestor until the mid 1700s (and our connection is unknown via traditional research), our two different tests should reveal couple of testing result differences creating a new branch in the genealogical time frame.

However, I still face the challenge of needing a third FGC test under L226 to eliminate all those pesky duplicate L226 YSNPs that are present but untested. I either sponsor an upgrade of a Big Y to FGC test or test these YSNPs individually by somebody very far up the L226 haplotree. Around 1/3 of my private YSNPs have been shown to be L226 equivalents from around 1,400 years ago (I guess the Dark Ages were unkind to L226 which almost did not survive). Testing the earliest branch of L226+ and FGC5660- (2 % of L226), will be my test for L226 equivalents. If this remote tester has positive results, they are probably L226 equivalents. If they test negative, they are either more private YSNPs to test or they could reveal another major branch under L226 that would split up our major L226 branching bottlenecks. If several in our cluster negative for these new private YSNPs, then we move up the haplotree to determine if they are major new branches.

Genealogical testing of YSNPs has arrived for me as well as several branches under L226 that are dominated by only one surname.

Update - Four more individual tests of private YSNPs has created another new L226 branch. FGC5647 is now the son of FGC5659 and the father of the branch FGC5639. So I am now L226>FGC5660>FGC5628>FGC5659>FGC5647>FGC5639. The two new branches were discovered by individual YSNP testing at a small fraction of the cost of one Big Y test.

lgmayka
04-27-2016, 02:12 PM
Most testers appear to be looking for a connection in the 1700 to 1800 AD range. How will those going with the absolute lowest price to fill out those older branches test the remaining 7,000,000 bases left totally out of their results when it comes time to move into modern times?
Some percentage (who can afford it) will order a higher-quality NGS test, to discover the needed SNPs. The rest will then order those SNPs individually.

But my main point was that most of the world's population has not reached that stage yet--they are still sorting out all the expansions that occurred before 1200 A.D.

FGC Corp
04-27-2016, 02:44 PM
I only got heavily into genetic testing since I have been a hard core genealogist for forty years now. My Casey research has hit the brick wall in western South Carolina in the mid 1700s. We know there are at least 40 different males lines in this time frame where little progress has been made over the last 40 years and speculation continues to turn into compiled family histories. With only YSTRs, the progress has been pretty minor but important: 1) We have confirmed that around twenty of these lines are indeed related but did find one outlier Casey living in the same area; 2) We are very fortunate to have very isolated genetic cluster with seven mutations since L226 creation which is only 1,400 years old (my ancestors rolled the dice and got 9s to 10s for long period of time). Our unique signature under the very isolated L226 signature can be assigned to our surname of one Casey progenitor. 3) Along the way, we have picked up two interesting NPEs to add to our analysis (one NPE is supported by probate documentation). 4) we discovered one major branch of a very rare marker value that divides the entire cluster into two 50/50 branches (this is pretty rare and lucky). 5) We have discovered one fairly close match that partially matches our signature but is obviously more distantly related to our cluster (its surname is Kersey which is a close match to Casey). The Kersey tester has a pending test of four private YSNPs that tested positive by others in our cluster.

But testing YSTRs alone is not making as much progress as anticipated by this community of researchers which have been exchanging traditional research for several decades and a lot of the researchers/sponsors are aging. Enter FGS tests. We now have our YSNP testing down to our surname cluster with the discovery of the new L226 branch FGC5639 which has only tested positive within our cluster of testers (after 48 Big Y tests and my one FGC test). Only one $39 individual test discovered this branch where 48 NGS tests failed to discover this branch. But it has been frustrating to get the L226 community to take any FGC tests which has two major advantages: 1) one for genealogical clusters, only FGC tests can produce the maximum number of private YSNPs to break down brick walls after surname creation (around 1,000 years ago for my Irish clan). 2) There are major bottlenecks where large numbers of private YSNPs are stuck under major branches. With the 24 L226 branches discovered to date, around one-third of all private YSNPs are attached to only three branches of the 24 known branches. Again, only higher resolution tests will break up these bottlenecks since 48 subsequent Big Y tests have failed to discover even one new branch in the trunk of the L226 haplotree that contains all three bottlenecks.

I convinced the most remote Casey line from my FGC Casey tested line to take the FGC test to discover more private YSNPs. He will be the only NGS tester to have show results on 12 my additional private YSNPs that 48 Big Y tests have not tested. Since his test is Y Elite 2.1 vs. my test which was Y Elite 1.0, he will hopefully find two or three additional private YSNPs with more coverage. Also, since we share no common ancestor until the mid 1700s (and our connection is unknown via traditional research), our two different tests should reveal couple of testing result differences creating a new branch in the genealogical time frame.

However, I still face the challenge of needing a third FGC test under L226 to eliminate all those pesky duplicate L226 YSNPs that are present but untested. I either sponsor an upgrade of a Big Y to FGC test or test these YSNPs individually by somebody very far up the L226 haplotree. Around 1/3 of my private YSNPs have been shown to be L226 equivalents from around 1,400 years ago (I guess the Dark Ages were unkind to L226 which almost did not survive). Testing the earliest branch of L226+ and FGC5660- (2 % of L226), will be my test for L226 equivalents. If this remote tester has positive results, they are probably L226 equivalents. If they test negative, they are either more private YSNPs to test or they could reveal another major branch under L226 that would split up our major L226 branching bottlenecks. If several in our cluster negative for these new private YSNPs, then we move up the haplotree to determine if they are major new branches.

Genealogical testing of YSNPs has arrived for me as well as several branches under L226 that are dominated by only one surname.

We have a number of people who have used our test for recent genealogical work, ie lines since 1700. I can put you in touch with them to compare approaches. Secondly, we can do some additional R&D in those cases to identify more challenging SNPs that require special investigation, such as those on the X-chromosome homologous region.

TigerMW
04-27-2016, 05:15 PM
Your favorite project admins and volunteers are pushing the tree forward and FTDNA has a Y team including a fairly new assignee who has been very responsive. I don't have time to do the checks again but several months ago I actually compared ISOGG's tree to FTDNA's haplotree for R1b. FTDNA had many, many more branches. FTDNA is not relegated to NGS tests for their haplotree so it is bound to be larger than the NGS only trees like the YFull's. This is spotty by subclade, and where FTDNA has developed SNP Packs. That is much of the reason I've pushed hard on getting FTDNA to get out complete coverage of R1b subclades via their Packs. If it goes into Pack it is automatically evaluated for their haplotree. It's not there yet, but growing in leaps in bounds.
...

I just ran a check this morning of R1b.
FTDNA has about 1,811 distinct branches in their haplotree for R1b.
ISOGG has about 827 distinct branches in their tree for R1b.

No one is failing. Everybody is behind. This is the constant state now as the tsunami of Y SNPs becomes a permanent high tide.

FGC Corp
04-27-2016, 05:24 PM
We have a number of people who have used our test for recent genealogical work, ie lines since 1700. I can put you in touch with them to compare approaches. Secondly, we can do some additional R&D in those cases to identify more challenging SNPs that require special investigation, such as those on the X-chromosome homologous region.

Also, there is potential to use additional sequencing approaches, i.e. other than Illumina or Sanger sequencing, if that would be helpful in special cases like these.

TigerMW
04-27-2016, 05:30 PM
... We need as many NGS testers as possible, which in turn requires the lowest possible test price--i.e., the Big Y on sale.
...

This is my example of using Big Y and STR matching at FTDNA. I sat around for a long time with 67 STRs and a couple of matches that seemed to have nothing to do with my surname. Geographically, however, there was a connection with my surname and the fact that two of the three were residents of Wales. On the other hand, since my immigrant ancestor was from Ireland it looked like a very old connection. Their surnames were Morgan and Evans.

My immigrant ancestor is Edmund Walsh, b.c. 1820, Co. Kilkenny, Ireland

Over time, a new match came in at 67, but still not that close, GD=6, and with a different surname. FTDNA changed their GD calculations and he became GD=5. He was GD=4 from me at 37 STRs so he doesn't even show up on my 12, 25 and 37 STR matching screens.

I was lazy and didn't pay attention to him but he also upgraded to 111 STRs. Now this new match was a GD=6, which isn't bad at 111 STRs. Lo and behold, when I reached him via the matching screen he realized that he should not leave his MDKA field blank. He filled it in:

James Welsh, b.c. 1826, Kilkenny, Ireland

Now we were getting somewhere. We both did Big Y.

There is an SNP named ZW02 that only we share. I think it is about 300 years old, but could be 200 years old.

In the 67 STR match list, we see a couple of Jackman's at GDs of 4 and 5. I'd like to get them to 111 STRs too, but they are ZW04+, which the Welsh/Walsh folks above are too.

The family stories line up with the Jackman's as Cambro-Norman immigrants to Ireland from Wales.

Here are my FTDNA Y matching screens at 111 and 67 STRs with the contact names scratched out.

https://dl.dropboxusercontent.com/u/17907527/Example_FTDNA%20_YSTR_Matches.pdf

Did you notice that my haplogroup assignment is R-ZW02 as is Welsh's? Jackman has the R-ZW04. The following screenshot shows the haplotree with ZW02 below ZW04. I see have to request them to knock them down one level from Z17909. All three of these SNPs are available on FTDNA's Advanced Tests menu.

https://dl.dropboxusercontent.com/u/17907527/Example_FTDNA_Haplotree.pdf

Here are a couple of Big Y matching screenshots. FTDNA needs to make this easier to use and stop its up-side down view.

https://dl.dropboxusercontent.com/u/17907527/Example_FTDNA_BigY_matches.pdf

What do I need? More SNPs and STRs would be okay. I'm all for more test result data. They certainly couldn't hurt, but that's not what I need. I need more individuals that might be related to me to test. I'd love to a sweep through Co. Kilkenny, Ireland and South Wales. We do have a ZW SNP slightly older than these two in Geno 2 NG but I'd rather have a sweep through these geographies with a really low priced Big Y.

FGC Corp
04-27-2016, 05:36 PM
We have a number of people who have used our test for recent genealogical work, ie lines since 1700. I can put you in touch with them to compare approaches. Secondly, we can do some additional R&D in those cases to identify more challenging SNPs that require special investigation, such as those on the X-chromosome homologous region.

We've done some of this custom work for a project of about 100 people plus, as well as a few others in-process.

miiser
04-27-2016, 07:46 PM
Over time, a new match came in at 67, but still not that close, GD=6, and with a different surname. FTDNA changed their GD calculations and he became GD=5. He was GD=4 from me at 37 STRs so he doesn't even show up on my 12, 25 and 37 STR matching screens.

This comment serves more as an indictment of FTDNA's poor and arbitrary "match" algorithm than as an endorsement of either STR, SNP, or NGS testing. Genealogically relevant matches popping in and out of the match list as a function of the number of STRs tested is an example of how broken their software system is.


What do I need? More SNPs and STRs would be okay. I'm all for more test result data. They certainly couldn't hurt, but that's not what I need. I need more individuals that might be related to me to test. I'd love to a sweep through Co. Kilkenny, Ireland and South Wales. We do have a ZW SNP slightly older than these two in Geno 2 NG but I'd rather have a sweep through these geographies with a really low priced Big Y.

I've been preaching for years that testing a greater number of people, especially potential matches, is usually a more effective use of funds than continuously upgrading single individual's tests to deeper testing via FTDNA upgrades, SNP packs, and Big Y. But in the past you have vehemently opposed this argument against widespread Big Y testing without discrimination. So I'm glad to see you've finally reversed your position on this strategy. But it risks creating the impression that you tend to favor arguments that encourage FTDNA testing, even when it requires abandoning your historically established position.

Some surname lineages have little need for NGS. Nearly every surname lineage will benefit from a greater number of testers in the database, especially from homeland regions. NGS should be used for targeted testing in specific cases where finer branch fidelity will be genealogically informative. In cases where people are testing just out of curiosity regarding the nitty gritty details of branching without expectation of much or any genealogically relevant information, and are willing to spend hundreds of dollars to scratch this itch, they might as well do it right and go with FGC.

TigerMW
04-27-2016, 08:36 PM
This comment serves more as an indictment of FTDNA's poor and arbitrary "match" algorithm than as an endorsement of either STR, SNP, or NGS testing.
...
I disagree. My criticism is not of FTDNA's match algorithm. It is basic a Genetic Distance construct. I think the conclusion is very clear, that at least in haplogroups like R1b, 67 or more STRs are needed. 37 or less are not usually enough. In my case, 111 was really needed.

In any situation, SNP validation is normally needed for tree identification.

RobertCasey
04-27-2016, 08:45 PM
I've been preaching for years that testing a greater number of people, especially potential matches, is usually a more effective use of funds than continuously upgrading single individual's tests to deeper testing via FTDNA upgrades, SNP packs, and Big Y. But in the past you have vehemently opposed this argument against widespread Big Y testing without discrimination. So I'm glad to see you've finally reversed your position on this strategy. But it risks creating the impression that you tend to favor arguments that encourage FTDNA testing, even when it requires abandoning your historically established position.

Some surname lineages have little need for NGS. Nearly every surname lineage will benefit from a greater number of testers in the database, especially from homeland regions. NGS should be used for targeted testing in specific cases where finer branch fidelity will be genealogically informative. In cases where people are testing just out of curiosity regarding the nitty gritty details of branching without expectation of much or any genealogically relevant information, and are willing to spend hundreds of dollars to scratch this itch, they might as well do it right and go with FGC.

There is a big difference between - seeking out possible testing candidates and enticing them to join via success stories. From my previous genealogical research, I know that there are around 40 to 60 Casey lines from South Carolina that have not been connected to date. Our project now has around 15 of these lines well tested at 67 markers or higher - so we are getting fairly good representation. We get another new line every six months or so. I think the discovery on two genealogical YSNPs is getting a lot of interest - both appear to be Casey only mutations. I am hoping that success stories of recent rapid progress of both YSTRs and YSNPs will entice the other lines to join our testing via our recent progress by word of mouth.

Progress with only YSTR tests just does not get there. We have around 25 YSTR tests in our cluster and genealogists are just not that impressed with our YSTR results. We also are very lucky of being very genetically isolated - L226 is very isolated and the Casey cluster is also very genetically isolated. There is no issue of convergence but YSTRs only are just not connecting our ancestors together. We have also had minimal success with 111 marker upgrades as well (every tester seems to have unique mutations). We are now testing through the remaining seven private YSNPs with some pretty good results (two new branches in the last couple of months that within our surname cluster). I also have 12 more private YSNPs not tested by Big Y and second person from our cluster just ordered another FGC Y Elite 2.1 to discover several more private YSNPs which would keep us busy for some time.

We do have a new major brick wall that genetics have revealed. Our Casey line just seemed to appear from nowhere around 1700 with no matches between 1000 AD and 1700 AD - except for one excellent partial match - Kersey. This person just tested FGC5647+ and created a new older branch that is a father of our Casey only FGC5639 branch which was discovered in the recent past. Testing six private YSNPs for the next closest match - Carey - all came back negative. Seven private YSNPs combined with 111 markers just will no cut it either.

I do not how much we will be able to connect with testing of private YSNPs when our brick wall remains around 1700 unless we can discover some more testers like the Kersey tester who is a pretty good partial match to our Casey cluster signature. But I know for sure, YSTRs by themselves are not up for the task. Even with high resolution NGS tests and 111 marker tests will not get us there. I think that we might have to look at 400 YSTRs to finally get all of the connections necessary. With 400 YSTRs, we will need a lot more testers - then we can use the impressive GENTREE YSTR approach which could create some amazing connections by using a lot more faster mutating markers but requiring a lot more testers (with known genealogical connections) to sort out the mutations.

TigerMW
04-27-2016, 08:49 PM
...
I've been preaching for years that testing a greater number of people, especially potential matches, is usually a more effective use of funds than continuously upgrading single individual's tests to deeper testing via FTDNA upgrades, SNP packs, and Big Y. But in the past you have vehemently opposed this argument against widespread Big Y testing without discrimination. So I'm glad to see you've finally reversed your position on this strategy. But it risks creating the impression that you tend to favor arguments that encourage FTDNA testing, even when it requires abandoning your historically established position.

Some surname lineages have little need for NGS. Nearly every surname lineage will benefit from a greater number of testers in the database, especially from homeland regions. NGS should be used for targeted testing in specific cases where finer branch fidelity will be genealogically informative. In cases where people are testing just out of curiosity regarding the nitty gritty details of branching without expectation of much or any genealogically relevant information, and are willing to spend hundreds of dollars to scratch this itch, they might as well do it right and go with FGC.
I'm not following you on this. Either that or my post was unclear.

What position did I reverse? I think I've been in favor of NGS testing almost from the outset, and in all honesty folks like FGC Corp were convinced me. My original position (since late 2013) was that every individual should drive for two NGS results per their surname per cluster and then one NGS result from each adjacent cluster.

This is line #1 of the Y classic screen of the R1b project. It's been this way for quite a while.
"A general recommendation: 1) 111 STRs, 2a) Big Y SNP discovery tests - two per surname per cluster, or 2b) R1b-M343 Backbone SNP Pack"

There is not much room in a subgrouping title but my intent is that people do both 1 and 2 with two being 2a for those who can afford it and 2b for those who can't.

The background page has this in italics:
"If you have a true interest in genetic genealogy and breaking beyond the brickwalls of your genealogical records, please strongly consider both Big Y to determine your haplogroup down to a very recent timeframe and 111 STRs so that you can refine your close-in family and surname mutation history tree."

I recently became more aggressive in conversations and recommend two NGS results per every genealogically known family, ideally between the two most distant male cousins of the genealogically known MDKA.

This does not mean everyone can afford that strategy but it is a good strategy in my opinion so I recommend it, as always, as long you can afford it. I don't assume I know what other peoples' budgets are.

TigerMW
04-27-2016, 09:00 PM
... Even with high resolution NGS tests and 111 marker tests will not get us there. I think that we might have to look at 400 YSTRs to finally get all of the connections necessary. With 400 YSTRs, we will need a lot more testers - then we can use the impressive GENTREE YSTR approach which could create some amazing connections by using a lot more faster mutating markers but requiring a lot more testers (with known genealogical connections) to sort out the mutations.
I'm not familiar with GENTREE but I agree with the concept that we should use all of the data possible, including stable SNPs, questionable SNPs, indels and all of the STRs we can. I think good analytics software can sort out a best fit tree. All of the mutations will need to have some kind of mutation rate or score for this but with no calls and the like we need good statistical analysis to calculate the best fit with the confidence scores.

This takes me back to FTDNA's supposed Big Y enhancements that they say will apply to existing customers. I think it would be worthwhile if they would take their cut at extracting STRs and stick them into a database. This will cause some complex programming to have STR panel matching and Big Y STR matching, hopefully integrated somehow. Ideally you'd leverage both systems as we know that the NGS BAM file STR extractions have their extra vagaries. I read some article that FTDNA did mention this at their last conference.

RobertCasey
04-27-2016, 09:07 PM
I've recently became more aggressive in conversations and recommend two NGS results per every genealogically known family, ideally between the two most distant male cousins of the genealogically known MDKA.

I am afraid you are correct about more than two NGS tests per surname cluster as a possible option. Of course, a mixture of 10 % higher resolution NGS tests is another option. Hopefully, it will not take two per proven genealogy as this would require my surname cluster to test over 100 NGS tests. Another alternative to so many NGS tests could be the use of faster mutating YSTRs. I think it would be an easier sell to get people to test higher resolution YSTRs if they are substantially less than NGS tests. GENTREE used to have a YSTR test that really should had been our strategy. They had 20 slow mutating markers and 20 very fast mutating markers (CDY range). With 50 to 100 economical tests (having people test known relatives to sort out mutations), they produced extremely well connected charts better than we are even today. Unfortunately, this division of Sorensen was acquired by Ancestry.com and then was discarded. We really need to explore using the 400 YSTRs that we are getting already. Once the read length is high enough, all 111 markers will be there as well which eliminate the need for YSTR testing for those testing NGS. Of course, then we have the database issue but FTDNA could add these if enough focus was put into this approach. A hybrid of all three - many more NGS tests, higher resolution NGS tests and using part of the 400 YSTRs may be required.

ChrisR
04-27-2016, 09:23 PM
- Most clades across the human haplotree have not yet been extended down to 1200 A.D. For all these clades, Mikewww's suggestion is exactly correct: We need as many NGS testers as possible, which in turn requires the lowest possible test price--i.e., the Big Y on sale.
- A few clades, presumably mostly of British Isles and Ashkenazi Jewish ancestry, have already been extended down to 1200 A.D. and later. These clades need higher precision, whether through a higher-quality NGS or individual SNP tests.
This has been my observation for some of the major Ashkenazi Jewish clusters in J2 which have very unsatisfying results from BigY which was not able to split up to 6 different lineages into distinct haplogroups. The expansion time seems to be under what BigY can measure with SNPs and no other option then upgrading to higher coverage remains. I'm relieved that I'm not involved in convincing the same kit owners to another NGS test in such a short time frame. Irony is that FTDNA want's to be the leader for Jewish genetic genealogy but for this case you should hear some of the comments from the Jewish Y-lineages project admins... ;-)

RobertCasey
04-27-2016, 09:26 PM
I'm not familiar with GENTREE but I agree with the concept that we should use all of the data possible, including stable SNPs, questionable SNPs, indels and all of the STRs we can. I think good analytics software can sort out a best fit tree. All of the mutations will need to have some kind of mutation rate or score for this but with no calls and the like we need good statistical analysis to calculate the best fit with the confidence scores.
They presented at the second RootsTech when it was primarily a software developers conference. They took an atDNA approach to YSTRs. Each sponsor had to test five or ten known cousins each and there had to be around ten different related but unconnected lines to test. Their results were very far ahead of its time in both concept and results. I immediately ordered this test but my money was returned due to the acquisition. They also had extremely good tree building software that was able to quickly sort the results. Speaking some software that shows some real promise:

http://www.anthrogenica.com/showthread.php?6840-An-alternative-to-Fluxus-for-R1b-L21-only

This works great but has two major issues - it is very CPU extensive and takes 10 to 30 minutes for 100 testers. Also, there is a timeout issue that Dave Vance is working on. If the performance issue addressed and limit is raised to 400 or 500 testers, it would automate trees which would allow us to make better recommendations for NGS tests, SNP packs and private YSNP testing. It has really revealed (with minimal prep work) the structure of the L226 haplotree with very good accuracy (a few small glitches here and there).

miiser
04-27-2016, 10:00 PM
There is a big difference between - seeking out possible testing candidates and enticing them to join via success stories. From my previous genealogical research, I know that there are around 40 to 60 Casey lines from South Carolina that have not been connected to date. Our project now has around 15 of these lines well tested at 67 markers or higher - so we are getting fairly good representation. We get another new line every six months or so. I think the discovery on two genealogical YSNPs is getting a lot of interest - both appear to be Casey only mutations. I am hoping that success stories of recent rapid progress of both YSTRs and YSNPs will entice the other lines to join our testing via our recent progress by word of mouth.

Progress with only YSTR tests just does not get there. We have around 25 YSTR tests in our cluster and genealogists are just not that impressed with our YSTR results. We also are very lucky of being very genetically isolated - L226 is very isolated and the Casey cluster is also very genetically isolated. There is no issue of convergence but YSTRs only are just not connecting our ancestors together. We have also had minimal success with 111 marker upgrades as well (every tester seems to have unique mutations). We are now testing through the remaining seven private YSNPs with some pretty good results (two new branches in the last couple of months that within our surname cluster). I also have 12 more private YSNPs not tested by Big Y and second person from our cluster just ordered another FGC Y Elite 2.1 to discover several more private YSNPs which would keep us busy for some time.

We do have a new major brick wall that genetics have revealed. Our Casey line just seemed to appear from nowhere around 1700 with no matches between 1000 AD and 1700 AD - except for one excellent partial match - Kersey. This person just tested FGC5647+ and created a new older branch that is a father of our Casey only FGC5639 branch which was discovered in the recent past. Testing six private YSNPs for the next closest match - Carey - all came back negative. Seven private YSNPs combined with 111 markers just will no cut it either.

I do not how much we will be able to connect with testing of private YSNPs when our brick wall remains around 1700 unless we can discover some more testers like the Kersey tester who is a pretty good partial match to our Casey cluster signature. But I know for sure, YSTRs by themselves are not up for the task. Even with high resolution NGS tests and 111 marker tests will not get us there. I think that we might have to look at 400 YSTRs to finally get all of the connections necessary. With 400 YSTRs, we will need a lot more testers - then we can use the impressive GENTREE YSTR approach which could create some amazing connections by using a lot more faster mutating markers but requiring a lot more testers (with known genealogical connections) to sort out the mutations.

I don't disagree with you.

In my experience, there are some cases where test upgrades and NGS are beneficial or even necessary, and other cases where they're not.

In my surname project, about 2/3 the members are able to be grouped, with very high confidence, based on a 37 marker test alone. The remaining ~1/3 of project members with ambiguous STR signatures have needed either SNP testing or 67 marker upgrades to determine with confidence which lineage they belong in. No one in my project has yet needed NGS in order to understand their placement in the tree. If they want a more precise estimate of the MRCA to other members of their group, I encourage them to upgrade to 67 or 111 markers. Just about every group has had SNP testing of one of its members to confirm the group's position within the tree.

I suspect my own lineage of being the result of an NPE from a different surname lineage. The tree position of this other lineage is not certain, but it looks like it may possibly be an old intact lineage from ~1300 AD with a fairly large GD. So I have suggested that one of them do NGS so that the rest of the possible group can test for the discovered SNPs. I have also offered to sponsor YSEQ testing of SNPs, not yet recognized by FTDNA, that were discovered through NGS and Alex's tree to confirm the group's tree placement.

In one lineage, we've got a surname group with an MRCA ~1700. The next closest matches are to a group of about 20 guys of various surnames, all from the same region, with an MRCA of around ~800 AD. What we really need to understand the history of this lineage is intermediate matches with MRCAs in the range of 1000-1600 AD, assuming such lineages are even extant. Hypothetically speaking, doing NGS of this entire group just to find out that Chapman, Curley, McMahon, etc. of central Ireland have an MRCA at ~1000 AD and Flood, Burke, and Dolan of central Ireland have a different MRCA at ~1000 AD - it doesn't really give you any useful information about the history. This is a case where I don't recommend NGS because it isn't expected to provide much information of value. This portion of the tree is too empty to benefit from deeper testing. What we need is more testers to populate this portion of the tree, and then later we can worry about NGS if it's needed. We're not there yet.

In theory, if all the members of each surname group in my project were to do NGS, they might be able to firmly establish the phylogeny of individuals within the group. But this would not be genealogically informative at all. The documentation in Ireland typically dead ends at about 1850. So, for example, it doesn't get you anywhere to find out that you have a ~1700 MRCA with Thomas born ~1850 in Roscommon versus a ~1600 MRCA with Michael born ~1850 in Roscommon.

Regarding publicity, word of mouth, and growth of the project: I've put a large amount of effort into recruiting via genealogy forums, Facebook, private messaging people via their trees, etc. I've had some success, with a burst of project joins in response to the effort. But there seems to be a limit at which the market has reached its potential. There are only so many people out there with an interest in genealogy who are willing to go to the trouble of doing a DNA test, and after exhausting the pool the project joins return to a slow but continuous trickle. But I doubt that most surname projects have exhausted these channels as thoroughly as our surname project has, and most will benefit from the increased testing generated by publicity.

I expect that at some point in the future (5 or 10 years?), WGS price will fall to the point that there's no good reason to do anything else. I think medical testing will be the driver of this. It's already common practice for doctors to order blood tests that cost a few hundred dollars, so it's not hard to imagine just about everyone getting a WGS test in the not-too-distant future. And once WGS for medical purposes is commonplace, genealogy will be an incidental beneficiary.

I worry that if we continue this discussion, a moderator may step in and say that we've meandered too far from the thread topic (specifically, Big Y vs Full Genomes) and these comments risk being exported to a new thread. So, having had my say, I'll leave it at this.

miiser
04-28-2016, 02:40 AM
There is a big difference between - seeking out possible testing candidates and enticing them to join via success stories. From my previous genealogical research, I know that there are around 40 to 60 Casey lines from South Carolina that have not been connected to date. Our project now has around 15 of these lines well tested at 67 markers or higher - so we are getting fairly good representation. We get another new line every six months or so. I think the discovery on two genealogical YSNPs is getting a lot of interest - both appear to be Casey only mutations. I am hoping that success stories of recent rapid progress of both YSTRs and YSNPs will entice the other lines to join our testing via our recent progress by word of mouth.

One follow-up response:

I just now realized that you seem to be suggesting an idea that I didn't immediately recognize - the idea that NGS testing of project members will lead to breakthroughs, and the excitement of these success stories will attract new testers into the project. So a broader database will be the natural outcome of NGS testing.

But I don't believe this excitement is the typical customer response of the shotgun approach for NGS testing (two per surname cluster, or whatever other general coverage strategy one favors). More often than not, I've seen people upgrade their test on the advice of an admin or forum recommendation, gain no new matches or genealogical information, and then ask, "Can someone please explain to me why I just threw away hundreds of dollars to have the same matches I already knew about, and an FTDNA Big Y matching list that my admin tells me I should ignore?" And the only lame answer they receive is, "Citizen science or something?" Anticlimactic let downs of unrealistic expectations are much more the most common result of advanced testing rather than exciting breakthroughs. This is why NGS testing needs to be focused on specific cases in which there is a specific question to be answered, and a realistic expectation that NGS is able to answer that question. Your distant match to Kersey is a good example of this sort of case.

IanFitzpatrick
04-28-2016, 11:37 AM
I think this post will be on topic because I am going to ask everyone which would be the best way to proceed with this situation Big Y or Full Genome

James Kane is familiar with what I am about to present.

Back in the fall I ventured into the realm of the BigY after talking to one of the Admins on the Fitzpatrick DNA project. He also told me one Fitzpatrick within was taking the BIGY. This group appeared through YSTR testing, to be closely related. The group called "the Nine of Clads" by Colleen Fitzpatrick, has been around for quite some time but has expanded, presently there are 30 possible Fitzpatrick's in the project that would be placed in the group by YSTR results alone

So, after I received my results and with the help and guidance of a great group of volunteers, I made it on to the Big Tree and after that within a few months, the branch ZZ44, has grown into 9 men, actually 10 men, one is not on the Big Tree.

Now, within that group there is a group of 4 men that have an unusual marker at DYS449=26. I looked into this and found it occurs in less than .02% of R1b men. Although it is relatively fast moving marker, it typically moves up in value not down.

To see what I am talking about just have a look down the thousands of L21 DYS449 markers and try to find one that equals 26

I also joined the Kirkpatrick name Project and was placed in a group. This group was made up of Fitzpatricks along with a few other surnames. About 25 men in total in the group and about 18 have DYS449=26, the rest have DYS449=27 which is also not common.

I never see much in the way of YSTR matches since I also have along with the DYS449 marker rarity, two other markers that are both under .01% of R1b men.

So, I think there is a cluster under ZZ44 that shares this marker mutation and it is specific to this cluster of men out of Cavan County.

I would like to hear what everyone thinks would be the best next steps, that is, would it be best to try to encourage a number of men in this cluster, which right now looks to be about 50 in total, to take a single SNP test, which would be cheap or encourage full testing?

I think if enough men took a single SNP test and showed that indeed there is a close relationship between the men with DYS449=26, it would encourage men to take a full test.

Big Y or Full Genome?

TigerMW
04-28-2016, 11:58 AM
Ian, I think we could have do each case one by one and the answers would vary, depending on the information and the framing of the real or hypothetical question, monies available, etc. In other words, the real answer is always "it depends" and in the final analysis that is always an individual decision for the payer.

What is your kit#? You are talking about STRs. I've got a very large database of haplotypes for L21 people. This does trend off topic but if you are looking at real individual cases we should look at all of the data. The R1b-L21-project yahoo group is set up as a Q&A for individual cases like this.

IanFitzpatrick
04-28-2016, 12:54 PM
Mike,

Thanks for your input, I agree with what you are saying but here is the situation that I believe more and more "amateur" genetic genealogists find themselves in.

As you even stated in your example, if things were going to move forward it meant you had to take the "bull by the horns" so to speak, to further your interest in your specific group, others here have also given the same examples. This is the place I find myself in. I have posted on the L21 group, but not unexpectedly, nobody really took any interest in this case. Not that I took that the wrong way, why would anyone really be interested in a group not their own?

FTDNA has not created a place for people to come together and further specific research beyond the project pages. What that would look like I have not really given too much thought but this place only exists now to a degree, within the Big Tree and projects.

Lets be honest, the L21 project is HUGE and one small cluster is not really on anyone's agenda. Have a look yourself for any DYS449=26, if there are any the are most likely already under ZZ44.

I have talked about this case with project Admins and received advice but if I want to further this I will have to use Social Media .....etc and hope to peak some interest. I do not really see me convincing anyone into spending the big bucks and I am not even sure I would want to try to convince someone to do this. But, a cheap method would be to try to get people to test one specific SNP and go from there?

As this field grows, as more and more people are looking for help, the existing volunteer support structure will not be able to properly deal with the needs or wants of individual who has little to no expertise. I have learned a great deal on this topic and know much more than I did a few months ago, but really who has the time and resources to do this, not too many.

Right now, the Big Tree is the best of the best resources that I can see out there, without this I feel I would have spent money for nothing.

I am the Fitzpatrick 332685 here

http://www.ytree.net/DisplayTree.php?blockID=1084

TigerMW
04-28-2016, 01:01 PM
... I think that we might have to look at 400 YSTRs to finally get all of the connections necessary. .. .
I will pursue this. I just hope I'm not too late in influencing that that they need integration of safe STRs into the regular STR panel database.

TigerMW
04-28-2016, 01:28 PM
Mike,

Thanks for your input, I agree with what you are saying but here is the situation that I believe more and more "amateur" genetic genealogists find themselves in.

As you even stated in your example, if things were going to move forward it meant you had to take the "bull by the horns" so to speak, to further your interest in your specific group, others here have also given the same examples. This is the place I find myself in. I have posted on the L21 group, but not unexpectedly, nobody really took any interest in this case. Not that I took that the wrong way, why would anyone really be interested in a group not their own?...
....
I am the Fitzpatrick 332685 here

http://www.ytree.net/DisplayTree.php?blockID=1084

I try to answer requests that are specific to kit #s of L21+ in the R1b-L21-project yahoo group. I'm sorry if I missed your inquiry but please never hesitate to repeat an inquiry or wack me up side of the head as needed.

I did a search of that yahoo group archives for "332685" but nothing came up.

When looking at STRs, I think looking at patterns is more useful in general than single STRs. DYS449 is fairly fast moving. I see 60 people in the R1b-L21_Hapotypes file with 449=26. There are handful at 25 too, but of course the modal is 29.

There is an unmistakable STR signature present that I've named 11134-z1714. The off-modal STRs are 389ii-i=17 534=14 and oftentimes 459=9,9 449<=27 511=11.

This is an important step. You want to identify your team. These (37 below) are the people who can help develop your section of the tree. Below are the 11134-z1714 people with 67 STRs that I've identified in the spreadsheet with their GDs at 67 STRs as the last digit(s). GDs of 18 and 19 are not close at all but the STR signature is still distinctive and FGC11134+ status seems to hold so we are probably looking at some serious branching if all if these people would do Big Y.

f332685 Fitzpatrick R1b-R1b-L21>DF13>FGC11134 0 (A)
f404921 Cratin zzL21suspect 8 (A)
f255916 zzzUnknown zzL21suspect 8 (A)
fH1638 Fitzpatrick zzL21suspect 9 (A)
fB6918 Kirkpatrick R1b-L21>DF13 9 (A)
f91296 Pugh zzL21suspect 9 (A)
f146389 Reilly R1b-L21 9 (A)
f262780 FitzClarence R1b-L21 10 (A)
f133193 Roark R1b-L21 10
f275360 Tracey zzL21suspect 10
f134287 Craton zzL21suspect 11 (A)
f635 McTernan zzL21suspect 11
f33464 Rahilly zzL21suspect 11 (A)
f31846 Lennon zzL21suspect 12
f116491 MacDonald zzL21suspect 12
f230226 Smith R1b-L21>DF13>FGC11134 12
f187737 Kelley R1b-L21 13
f200639 Kelly zzL21suspect 13
f20797 Mahon zzL21suspect 13
f227092 McConnell zzL21suspect 13
f176224 Ferrel R1b-L21>DF13>FGC11134 14
f91838 Gilroy R1b-L21 14
f158521 Lynch R1b-L21 14
f213150 McConnell zzL21suspect 14
f107579 Copeland zzL21suspect 15
fN4964 Gilgunn R1b-L21>DF13>FGC11134 15
f145148 McHugh R1b-L21>DF13>FGC11134 15
f17137 Kenny R1b-L21>DF13>FGC11134 16
yEBJTD McHugh zzL21suspect 16
f355688 Easton R1b-L21 17
fN5039 Hannaway R1b-L21 17
f251030 Lennan R1b-L21>DF13>FGC11134 18
fN71588 McDermott R1b-L21 18
f118718 McGarity zzL21suspect 18
f252763 zzzUnk(Wood) R1b-L21 18
fN3365 Darby R1b-L21>DF13>FGC11134 19
f18697 Mathis R1b-L21 19

Everyone from Rahilly above marked with A also have off modals 389i=12 AND 413b=24 exclusively for this group. That's your close in team. Think of them as your "A" team. In fact, this is clear enough I'll create this subset as variety 11134-z1714-A in the spreadsheet.

Anyway the net is these people are already
1) in the FTDNA system (with DNA in storage and some way to contact them)
2) have some inclination for advanced testing (they are at 67 or more),
3) are the people you care about investigating.

You could try to get them to test for your novel Fitzpatrick SNPs discovered in Big Y and documented by Alex and James or whoever. From my perspective, to be fair to the people on your team, we should recommend Big Y so they will have their own lines of SNPs even if their connections to your novel Fitzpatrick SNPs are ancient (as some will be ... they will just be public SNPs as the branching unfolds.)



Right now, the Big Tree is the best of the best resources that I can see out there, without this I feel I would have spent money for nothing.
I absolutely agree. I've posted elsewhere that Big Tree is like a national treasure for P312 people. I'd just like to see the DF19 and U152 share more with the process.

I do have my own process (that I just do for L513) similar to Big Tree but it is generally manually based. Folks like James and Robert probably due too, but Big Tree is nice and easy for everyone.

IanFitzpatrick
04-28-2016, 02:11 PM
Mike, that is very helpful and much appreciated!

This I guess should be taken off line but remember, FTDNA put me and still today has me at L21 - YFull put me at FGC11134 - Big Tree put me where you see with a strong surname match and a tree to begin my genealogical search.

But this is my point, that obviously did not take much time to compile. Yes I understand that you have put in a tremendous amount of work into a compiling valuable dataset, but is this not the type of information that the tester should be providing as a result of the test itself. What is frustrating is you came up with this list, Alex, James and many other have steered me in the right direction and FTDNA is the one collecting the money.

That being said, with all this information in hand, what would be the best test? Big Y or Full Genome? Is the fact it can be put in a FTDNA project page that important?

Off Topic

In the Fitzpatrick DNA study there are 30 men with this marker, what are the odds of this group have this rare marker and the same surname? I feel this needs to be explored. Some people would be happy just knowing they match to this group without spending a lot of money. This would or could help in matching people as a group, on the cheap, so research by paper could be possibly given a certain level of credence.

Example, my MRKA is from Dublin about 1799, he was a Silk Weaver. I have a suspected father, he was a Flax Weaver according to records about 1792 in Dublin. Cavan County was a center of Flax weaving so this would be my link to Cavan and possible other Fitzpatricks listed as Flax weavers in Cavan.

The group of 4 men under 7716880-A-G and one other that has done the BIG Y all have DYS449=26

It may be fast moving but the odds are astronomical this marker has not been passed down through this subclade and the Big Tree shows it may have been stable at this value this way for quite some time.

According to YFULL
STR: DYS449 Detected in 77.29% samples
For R1b only
For all HGs
26 0.15%
27 2.10%
28 12.16%
29 39.94%
30 34.68%
31 7.81%
32 2.85%
33 0.15%
34 0.15%

TigerMW
04-28-2016, 02:48 PM
...
But this is my point, that obviously did not take much time to compile. Yes I understand that you have put in a tremendous amount of work into a compiling valuable dataset, but is this not the type of information that the tester should be providing as a result of the test itself. What is frustrating is you came up with this list, Alex, James and many other have steered me in the right direction and FTDNA is the one collecting the money.
If this is why you are posting here then this really is a different topic that is better on the FTDNA category of Anthrogenica or even on the FTDNA forum itself in the "Gripes and Grumbles" section.

My only comment is there are trade-offs. I would not FTDNA advise to beef up their analysis services. I'd much rather have them enhance function, usuability and performance of their I/T systems while producing tests like Big Y at a lower cost. My experience is that many of us (James, Robert, Alex and Vince are great examples) have much stronger professional capabilities and probability get paid more in our fields. We don't want FTDNA hiring people like us if that causes them increase their prices.

It is pretty standard in many industries that there primary providers of the bulk of the solutions but there are business partners and consultants that provide ancillary services. The niche marketeers may specialize in certain functions, geographies or vertical markets. I don't see why it should be any different in this industry.

IanFitzpatrick
04-28-2016, 03:00 PM
No, I am not complaining, I think you have answered my question and I agree with your thoughts


It is pretty standard in many industries that there primary providers of the bulk of the solutions but there are business partners and consultants that provide ancillary services. The niche marketeers may specialize in certain functions, geographies or vertical markets. I don't see why it should be any different in this industry.

If this is indeed the best solution then as a testing provider only, what is the best choice, Big Y or Full Genome?

My only comment is that FTDNA is trying to be more than just a testing company and not doing a very good job at it, they should listen to the experts.

TigerMW
04-28-2016, 03:11 PM
... Is the fact it can be put in a FTDNA project page that important?
Yes but its not just the project pages. This is particularly important if we really ever want to get to the day where we are not dependent on the passion, volunteerism and spotty (by subclade) leadership of administrators and other advocates.

This is the morbid truth. We all come and we all go.

Why do we do this? I suggest it is for posterity. We want want our descendants to know who we are/were and our ancestors were. We want to find relatives we didn't know about it and for them to find us if even if we are no longer here. We are passing on our knowledge of our families.

FTDNA has the best chance at keeping on-going, self-sustaining project, matching and sharing system. They have secession plans and full time employees who need FTDNA to be successful. They have to worry about legal things and have too much going that they can't just go bankrupt and leave us high and dry.

FTDNA's I/T systems have come along way and they keep progressing. The provide a platform that is broad and supports autosomal and mtDNA as well as multiple types of Y DNA testing.

FTDNA acquires and grows their own customer set, which is vital to the growth of our matching databases - important to us as individuals as well as to the niche competitors and business partners. The list I posted of your "team" prior was not magic. Those are people FTDNA acquired in some manner.

The more beyond the project page support is what I posted earlier as screenshots for myself that included SNPs of 200 to 800 years old.

Here are my FTDNA Y matching screens at 111 and 67 STRs with the contact names scratched out.

https://dl.dropboxusercontent.com/u/17907527/Example_FTDNA%20_YSTR_Matches.pdf

Here is where I sit in the haplotree, which turns out to be a much, much larger haplotree than ISOGG's.

https://dl.dropboxusercontent.com/u/17907527/Example_FTDNA_Haplotree.pdf

Here are a couple of Big Y matching screenshots.

https://dl.dropboxusercontent.com/u/17907527/Example_FTDNA_BigY_matches.pdf

FTDNA needs to make these things better so we don't have to depend on the spreadsheets and highly technical people.

By the way, did you and your compatriot Fitzpatrick do Big Y tests? If so, you will see a haplogroup label change too. It may take a couple of months but it will happen. Hopefully, new matches that FTDNA acquires will stumble upon you and your new Fitzpatrick SNPs, inquire and test for them.

I would be very surprised if your "A" team of matches is all encompassing. There could be other surnames with specific origins or better yet people from those locations that surface in the future.

TigerMW
04-28-2016, 03:35 PM
...
My only comment is that FTDNA is trying to be more than just a testing company and not doing a very good job at it, they should listen to the experts.
I agree they should listen to the experts. They do more than you know but there are many, many experts passionate about their concerns so it is a wait in line deal. Still, I don't want them to hire in a way to increase the costs of their tests.

Ironically, you bring up a different point - FTDNA is a testing company. They have their own plant, equipment and storage facilities. Some competitors are not full product line test providers and some aren't even testing laboratory companies at all, and are really just marketing agents and interpretation services providers.

We've heard about business economics earlier in this thread but we may we see a direct trade-off here.

An actual testing company with its own plant and equipment may be able reduce costs more than we think by their investments and the scaling of greater utilization of fixed equipment capacity and negotiations with component requirements, lowering per unit costs.

The trade-off of the greater investment and lower pricing might be reduced agility. A marketing agent and interpretation service provider could switch (think outsourcing) to a new testing company quite quickly.

Neither way is bad nor good. Both have their advantages and its good for consumers to have choices.

IanFitzpatrick
04-28-2016, 03:45 PM
I would be very surprised if your "A" team of matches is all encompassing. There could be other surnames with specific origins or better yet people from those locations that surface in the future.

This is exactly why I am looking at what is the best next step. My one and only significant YSTR match from FTDNA is at 67 markers with a GD of 4 and he does not know who his paternal Grandfather is.

Yes, from the Big Tree, the other Fitzpatrick that matches me on 91 of 108 novels is one of the 1300 matches I have on Big Y. He is listed as FGC11134 so how could I not also be FGC11134?

Even though I agree and understand your points made in the last post, my opinion is there is no value in the services FTDNA provides beyond the test itself at this time, so I am still not clear on best next steps

TigerMW
04-28-2016, 04:16 PM
Ian, I respectfully disagree. FTDNA provides much more than just a physical test result. We've already gone through a litany of services including DNA storage, public and private projects (including surname, geographic and by test type,) matching member communications while maintaining privacy, actual family trees, and - yes - test interpretations.
They are just too conservative on their test interpretations and they are way behind on any phylogenetic analysis - leaving us to do that to get our subclade ahead of the next guy's in line. They don't proactively make manual adjustments and deep diagnostics like many do, but you could have your haplogroup updated if you choose to request it. You could make the request and solve that problem quickly. It will be a moot point soon anyway as you'll have downstream SNPs marking your new terminal haplogroup.

IanFitzpatrick
04-28-2016, 04:35 PM
Oh I did ask and this was the reply Oct 10, 2015


Dear Ian,
Thank you for your response. Unfortunately, we cannot change your results based on the analysis of an outside party. Your Big Y test did not yield confident results for some of the SNPs downstream of L21. Thus the furthest we can define without further testing is L21 unless other SNPs are added to tree. I apologize for the inconvenience.

and look how far I got without FTDNA and believe that if I had gone elsewhere would most likely ended up in the same place so really the original poster was asking which way they should go and I hope this discussion has given them at least some insight into the choice they have.

What services a company supplies and the quality of service are two different things, Microsoft has OneDrive Cloud and it pales in comparison to Dropbox.

We can definitely agree to disagree, I think we are looking at things through different lenses, I am looking at this through the lens of the uneducated consumer that I am.

TigerMW
04-28-2016, 04:46 PM
Oh I did ask and this was the reply Oct 10, 2015
...
I guess I was right in saying they were too conservative, at least in your opinion. As I noted earlier, "It will be a moot point soon anyway as you'll have downstream SNPs marking your new terminal haplogroup."

You made a wise move having both you and your compatriot get Big Y. You've found a new line of SNPs that form a new branch, no doubt about it. Given the FTDNA Y STR results, also have about three dozen other people you could work with to further refine your Fitzpatrick line of SNPs. The headache is in getting responsiveness out of the other people on your "A" team but least you have the FTDNA provided platform to work on. The other members of your "A" team probably already have DNA on storage at FTDNA and just need to press a few buttons if you can convince them to test. The absolute price will be as low as you can get when another sales come up. It won't do you much good if you get a new person to do a test with greater coverage since the same locations won't have been tested on your branch forming Fitzpatricks. Still, it always yours and their choice on how to spend money.

We are glad to have you on board with the Anthrogenica forum. You don't have to answer, but I'm just curious. How did you hear about this thread?

FGC Corp
04-28-2016, 05:29 PM
Oh I did ask and this was the reply Oct 10, 2015



and look how far I got without FTDNA and believe that if I had gone elsewhere would most likely ended up in the same place so really the original poster was asking which way they show go and I hope this discussion has given them at least some insight into the choice they have.

What services a company supplies and the quality of service are two different things, Microsoft has OneDrive Cloud and it pales in comparison to Dropbox.

You can also get second opinions via a variety of R1b - U106 admins, in particular Iain MacDonald has done a very sophisticated analysis. He gets into the technical aspects and tradeoffs (and he is not affiliated with us in any respect). That analysis includes coverage and analytical analysis. He also happens to have a PhD in astrophysics.

IanFitzpatrick
04-28-2016, 05:32 PM
Completely understand that patience is the order of the day and this I know having researched my tree for about 5 years now and seeing how long it can take to squeeze out a breakthrough. I guess what appears to be a complaint is more along the lines of frustration that it could be so much better. However, I say this knowing that FTDNA has a number of thing to deal with and first and foremost it has to make money to survive as a business.

I have been reading this board for many months now and have learned a great deal from this resource, it is a diamond in the rough in my eyes and could also be a very useful tool to bring researchers together. The format is much better than the Yahoo group which I find is difficult to follow without being overwhelmed by too many emails.

This thread caught my eye since I am looking for a way forward in progressing what I see a really interesting branch on the genetic tree, in my eyes anyway lol. I also have a Fitzpatrick on my maternal side from Cavan County.

TigerMW
04-29-2016, 01:43 AM
....
I have been reading this board for many months now and have learned a great deal from this resource, it is a diamond in the rough in my eyes and could also be a very useful tool to bring researchers together. The format is much better than the Yahoo group which I find is difficult to follow without being overwhelmed by too many emails.
...
The reason I asked was that all of your posts so far have been on this thread. Something here definitely tripped your trigger. As an R1b-L21 person, I invite you to look at the R1b sections, particular the R1b-L21 categories.

TigerMW
04-29-2016, 02:01 AM
I checked the new YFull tree 4.4 out as well.

Number of distinct branches in R1b:

FTDNA = ~1,811

ISOGG = ~827

YFull = ~789


I just ran a check this morning of R1b.
FTDNA has about 1,811 distinct branches in their haplotree for R1b.
ISOGG has about 827 distinct branches in their tree for R1b.

No one is failing. Everybody is behind. This is the constant state now as the tsunami of Y SNPs becomes a permanent high tide.

RobertCasey
04-29-2016, 03:27 AM
I checked the new YFull tree 4.4 out as well.

Number of distinct branches in R1b:

FTDNA = ~1,811

ISOGG = ~827

YFull = ~789

I find that around 5 to 10 % of FTDNA's terminal YSNPs are based
on equivalent YSNPs in their own haplotree. So user beware when
looking at terminal YSNPs in their YSTR and YSNP reports. There
are also terminal YSNPs routinely do not match their haplotree -
but the haplotree is so vastly improved, I welcome a few glitches
here and there.

I also find that FTDNA new trees have 30 to 40 % more branches
than even Alex's charts which is making some serious progress.
Of course, FTDNA has exclusive access to all Big Ys but poor Alex
and others beg for copies from this database one test at a time.

It would be really nice if FTDNA had a default share BAM file option
and they offered full access to BAM files. They should charge $100
per month for this service to limit usage to 10 or 20 people who
could coordinate the analysis. These charges would be necessary
to offset the development effort and IT operational costs. But this
is unlikely to happen as this data is an asset to FTDNA and they
know that allowing better access would result in more intelligent
ordering (lower revenue/profits). Plus this requires more IT strength
to implement as well and like many companies, FTDNA views IT
as an expense vs. a marketing advantage of keeping ahead
of the competition.

TigerMW
04-29-2016, 03:34 AM
Well, this is interesting. A lot of people, including myself, have a lot of respect for Thomas Krahn's knowledge.

I guess this is his answer. Do NOT get FGC Elite Y Next Generation Sequencing. Do NOT get Big Y.

"Today you can get a whole genome sequence from Full Genomes Corp. or elsewhere below $1000. That includes a nicer Y coverage and longer reads than the biggie test from FTDNA. Forget about Big Y (and Elite) entirely. Go for WGS immediately."
https://groups.yahoo.com/neo/groups/R1b1c_U106-S21/conversations/messages/42643

I emboldened his key recommendation that is ironic.

I think he is a very knowledgable technician and good friend of FGC as well as being the ex-Lab Director at FTDNA.

However, I'm not sure if pricing, matching and general convenience and support considerations are always a part of the way a type typical technical assessment is made.

I still don't get why some don't see the simplicity and reality of genetic genealogy. You need as many people as possible to test at the same locations by the same measurements so you can do consistent comparisons. This has been called the Relative Comparison Method.

There is nothing wrong with experimenting with new tests and gathering more kinds of data, but for the genetic genealogy the power of more individuals tested is paramount which means lower pricing is needed.

I kind of think FGC Corp (the prinicipal and poster) agrees neither Big nor FGC Y Elite are the best choices. This may well be part of the FGC strategy of creating a niche for WGS. I think this makes perfect sense, business-wise. However, for genetic genealogists the boring reality may just be we need more people which means lower prices and not necessarily innovative new tests. We don't necessarily need more coverage, we need cheaper more than we need more.

I guess I should go out and buy the biggest, baddest new iphone I can get today, even if it only works in China or wherever.:)

TigerMW
04-29-2016, 03:43 AM
I find that around 5 to 10 % of FTDNA's terminal YSNPs are based
on equivalent YSNPs in their own haplotree.
I counted distinct branches, Robert. I did not double count equivalents from any of the trees. I'm not sure how you came up with your calculation. I have most of the R1b SNP packs position-anc-der detail and references to the various draft trees - across the board, not just for L226. The SNP pack development process is what has spurred their haplotree updates. There is no downplaying it.
They have way more distinctive branches than everybody else and this is growing weekly. I think L226 will get a shot in the arm soon, just make sure you are giving your input to Dennis.

This should NOT be astounding news other than they were behind a long time so it has just taken them a while to get their train engine running and on the right track. This is business as usual for them. They are slow but they are a true self-sustaining operation and eventually get things working.



I also find that FTDNA new trees have 30 to 40 % more branches
than even Alex's charts which is making some serious progress.
Of course, FTDNA has exclusive access to all Big Ys but poor Alex
and others beg for copies from this database one test at a time.

Yes, I'm probably beggar #1 and have sent out over1800 emails requesting testers to download their raw results and send them to me or directly upload them to the yahoo groups or otherwise follow Big Tree instructions. I've also set up a shared administrative profile at FGC that helps. I guess I'm bragging a bit, but there is no magic in this. The good news for me that in many subclades, like L21, critical mass has been reached and the NGS discovery and analysis process occurs of its own volition.

I've made the offer to send out emails to the U152 folks too. I really just want to see R1b get full coverage. I have mostly R1b lineages, but have now found at least one R1a lineage (like good Czech should). I just can't wrap my head around that yet.

RobertCasey
04-29-2016, 03:57 AM
"Today you can get a whole genome sequence from Full Genomes Corp. or elsewhere below $1000. That includes a nicer Y coverage and longer reads than the biggie test from FTDNA. Forget about Big Y (and Elite) entirely. Go for WGS immediately."
https://groups.yahoo.com/neo/groups/R1b1c_U106-S21/conversations/messages/42643


Mike, you really know the problem with that approach. Neither FGC or YSEQ offer FTDNA compatible enough YSTR and YSNP reports, so they can be easily integrated with FTDNA data. Also, all those YSTRs are being wasted by both FGC and FTDNA NGS tests until FGC creates a database of YSTR and YSNP reports that is easy to digest and appear more compatible with FTDNA YSTR and YSNP reports - as well as our EXCEL spreadsheets where the real databases currently reside.

Also, none of the current offerings have read lengths long enough to catch all 111 markers (people would not go for losing only 5 or 10 YSTRs). And a third note, WGS tests are still higher cost than Y Elite that have the same amount YCHR coverage.

There is no doubt, that in the next year or two, Y enrichment will be dropped by FGC when WGS costs drops below existing Y Elite costs. Hopefully, read lengths will also be in the 250 to 500 range to be able read all 111 YSTRs with no errors.

TigerMW
04-29-2016, 04:15 AM
Mike, you really know the problem with that approach. Neither FGC or YSEQ offer FTDNA compatible enough YSTR and YSNP reports, so they can be easily integrated with FTDNA data. Also, all those YSTRs are being wasted by both FGC and FTDNA NGS tests until FGC creates a database of YSTR and YSNP reports that is easy to digest and appear more compatible with FTDNA YSTR and YSNP reports - as well as our EXCEL spreadsheets where the real databases currently reside.

Also, none of the current offerings have read lengths long enough to catch all 111 markers (people would not go for losing only 5 or 10 YSTRs). And a third note, WGS tests are still higher cost than Y Elite that have the same amount YCHR coverage.
I think I'm following you but that means you agree with me that Thomas Krahn's statements really aren't a good recommendation for genetic genealogists in general. However, Thomas might be right for prestigious (not my term) customers and academic research studies.


There is no doubt, that in the next year or two, Y enrichment will be dropped by FGC when WGS costs drops below existing Y Elite costs.
Does FGC Corp agree that Y Elite is at end of life as a product? It may well be, but I don't think they would want to say that.


Back on the compatibility issue, it is truly important. Forget between vendors, it's hard enough within a vendor. As an example, it took over a year for FTDNA to integrate its Advanced Tests (Sanger Sequencing), SNP Pack and Big Y (NGS) results into an integrated system. This causes problems, as we see from Iain, that FTDNA becomes very conservative with their NGS interpretations to try to provide "safe" integration. Now they are trying to integrate Geno 2 NG results and we know of the false positives problems with some SNPs in Geno 2 NG so the integration work is continual and extensive.

It might be a great benefit if FTDNA would extract "safe" STRs out of Big Y and complement the 111 STR panel set with those. I'm afraid they'll take the easy way out rather than true integration. We'll see, but it could be a great boon for existing Big Y customers.

RobertCasey
04-29-2016, 04:19 AM
I counted distinct branches, Robert. I did not double count equivalents from any of the trees. I'm not sure how you came up with your calculation. I have most of the R1b SNP packs position-anc-der detail and references to the various draft trees - across the board, not just for L226. The SNP pack development process is what has spurred their haplotree updates. There is no downplaying it.
They have way more distinctive branches than everybody else and this is growing weekly. I think L226 will get a shot in the arm soon, just make sure you are giving your input to Dennis.

This is a delta between their terminal YSNPs reported in their YSTR reports and their haplotree where you order YSNPs. If you look at the terminal YSNPs, there are dozens of equivalents listed as terminal YSNPs based on their own haplotree where you order YSNPs. It is not real bad - just a little annoying that these two are no in sync. Most of this could be that they allow equivalent YSNPs to be ordered and then do not map the equivalents to any master branch defining YSNP. Many times the terminal YSNPs are not the first YSNP in the list of the haplotree where you order YSNPs.

PS. Just sent FGC5647 to Dennis as a new branch today - but that train may have already left the station since the list has already been submitted. I just pulled all the YSEQ data again yesterday but did not see any new branches. BTW- branches being discovered by individual YSNP testing are still running at around 10 % of the cost per YSNP of Big Y costs per branch. If I can just get Dave Vance's wonderful tool working at higher numbers, analysis of private YSNPs would be much much easier. His tool is on the verge of being sliced bread for individual YSNP testing analysis (as well as more intelligent ordering of NGS tests and intelligent ordering of panel/pack tests).

TigerMW
04-29-2016, 04:26 AM
Well, this is interesting. A lot of people, including myself, have a lot of respect for Thomas Krahn's knowledge.

I guess this is his answer. Do NOT get FGC Elite Y Next Generation Sequencing. Do NOT get Big Y.

"Today you can get a whole genome sequence from Full Genomes Corp. or elsewhere below $1000. That includes a nicer Y coverage and longer reads than the biggie test from FTDNA. Forget about Big Y (and Elite) entirely. Go for WGS immediately."
https://groups.yahoo.com/neo/groups/R1b1c_U106-S21/conversations/messages/42643

Vince T posts here too so maybe he will comment as he did on the U106 form. It appears the all out press is on.

Vince T followed Thomas post with ,"Yup, the laggards that have been putting off Big-Y and Y-Elite have chosen wisely. They need to wait until WGS 30x comes down in price, or even longer once nano-pore (whole-molecule) sequencing becomes viable, cheap, and ultimately succeeds NGS in quality"

The implication is FGC Y-Elite is dead. Do existing Y-Elite customers get an upgrade price with full credit for their investment or a data migration/conversion?

Of course Thomas and Vince are implying Big Y is dead too, but FTDNA may not agree. There are no indications that they do. They have their own plant and equipment so I don't know how we know their real costs and capabilities. It might be wise to consider Sun Tzu's advice in the "Art of War" very carefully, “If you know neither the enemy nor yourself, you will succumb in every battle”.

Mac von Frankfurt
04-29-2016, 03:10 PM
When one purchases a technology product they follow the market and time their purchase to maximize value and half-life for their purposes. The old adage, look before you leap, has never been more true.

As it turns out I am happy with my employer-issued Blackberry. Most think I am mad...mad I say!

lgmayka
04-29-2016, 03:36 PM
Number of distinct branches in R1b:

FTDNA = ~1,811

ISOGG = ~827

YFull = ~789
I feel obliged to make two rather obvious points:

- The three haplotrees have somewhat differing sources of information, and differing philosophies of tree inclusion. The net result is that each has its own advantages and shortcomings.

- R1b is not the only haplogroup on earth. :) Other haplogroups show different patterns. ISOGG's haplotree for N (http://www.isogg.org/tree/ISOGG_HapgrpN.html) still shows P189.2 in the wrong place, several years after its discovery and placement proof.

Amerijoe
04-29-2016, 03:52 PM
Thanks for pointing that out. I thought R1a was being phased out. Scottish R1a lets unite and drive them from our shores. I apologize, a doppelgänger of Mel Gibson took over.:hug:

Joe B
04-29-2016, 04:35 PM
I checked the new YFull tree 4.4 out as well.

Number of distinct branches in R1b:

FTDNA = ~1,811

ISOGG = ~827

YFull = ~789
Is the FTDNA haplotree readily available to the public? Can it be used for citation in published papers?

TigerMW
04-29-2016, 04:41 PM
"Number of distinct branches in R1b:

FTDNA = ~1,811

ISOGG = ~827

YFull = ~789"


I feel obliged to make two rather obvious points:

- The three haplotrees have somewhat differing sources of information, and differing philosophies of tree inclusion. The net result is that each has its own advantages and shortcomings.

I agree and did not intend to imply anything different. That's what I was trying to say in reply #136.
"This should NOT be astounding news.."

Perhaps one of the most important aspects is not just that FTDNA has access to all of their BAM files but also that they there have been several thousand SNP packs and Sanger Sequencing tests, that all go in to the same database. (Not sure if this is technically a singular database though.) We are almost to 2,000 SNP Packs done for people in the R1b project. There has to be many more outside of the project. Many of these packs have phylogenetic equivalents included so additional refinement in branching is being discovered.

Both the ISOGG and FTDNA trees have one clear advantage. They are both purposed at providing a comprehensive picture of the tree. They do not relegate themselves to an NGS perspective. We still need the NGS perspective, particularly for tree discovery, but in terms of production systems and matching, we need the full (all sources) tree.


- R1b is not the only haplogroup on earth. :) Other haplogroups show different patterns. ISOGG's haplotree for N (http://www.isogg.org/tree/ISOGG_HapgrpN.html) still shows P189.2 in the wrong place, several years after its discovery and placement proof.
Agreed, I keep my R1b hat on so I can maintain some intelligence in terms of commentary. The FTDNA tree definitely has errors and many pending updates.

However, some day I'll have to get involved with R1a too. The Vaneks, Kulhaneks, Kovariks, Sestaks need me to get on board. I've got at least one confirmed R1a line now but there has to more.

JamesKane
04-29-2016, 05:38 PM
On the topic of Big Y and Y Elite being dead... Look at the cost delta of going with GenomeGuide versus Y Elite. This is about the cost of Family Finder. You lose a bit of Y coverage, but it's still more than most Big Ys. You gain the full mtDNA sequence and most likely every thing of note for the Family Finder. The autosomal results can then be formatted for upload into something like Gedmatch or pose as an Ancestry test into FTDNA.

That's why WGS is becoming interesting. The price isn't too much more than piecing your complete genome together with FTDNAs offerings. In some ways it's superior and others there is an inconvenience factor.

I don't see why FGC (or FTDNA) should be discounting or crediting back because something better is here now. My Y Elite 1.0 is still in the top 30 of 400ish R1b-P312 BAMs sent my way. For Y line genealogy those results won't need to be updated until something much better comes along.

TigerMW
04-29-2016, 05:39 PM
When one purchases a technology product they follow the market and time their purchase to maximize value and half-life for their purposes. The old adage, look before you leap, has never been more true.

As it turns out I am happy with my employer-issued Blackberry. Most think I am mad...mad I say!

I think there is another aspect of this, which I understand well but am disappointed that I've not articulated well.

These tests are not consumables. You don't order them and eat them and that's it and buy another when you are hungry again.

At least that is not what genetic genealogy customers want generally.

Vendors needs to provide migration paths which protect and leverage customer investments.

I remember back in 2013, I believe, some smart folks commented that we will have testing companies, test remarketing companies, test interpretation and consulting services companies, and test result web application & database companies.

Of course, there will be combinations of the above.

I'm using the word database loosely here, but when it comes to the genetic genealogy we don't all want to be our own systems integrators and technology migration managers. FTDNA provides this capability. Not particularly well or speedy, but they do provide something that just isn't out there. They provide data accumulation, integration and access along with upgradability and backward compatibility. I wouldn't be surprised if they don't spend more money on this than they do on testing equipment. It restricts their agility enormously because they want to minimize leaving current customers in the ruins.

I'm describing this capability poorly but it is vital. If all of your family genealogical records went up in flames you'd feel the importance of the concept pretty quickly. If you had a deceased relative who's DNA was no longer available you'd feel the pain of not being able to compare with new technology test results. Think of the problem as the new results come in another language.

If you bought a house for the rest of your life but found out it was worn out or had some major foundation fault and needed to be torn down and rebuilt you wouldn't feel good about that, even if they told you could do it all yourself with a few good tools.

lgmayka
04-29-2016, 08:18 PM
Both the ISOGG and FTDNA trees have one clear advantage. They are both purposed at providing a comprehensive picture of the tree. They do not relegate themselves to an NGS perspective. We still need the NGS perspective, particularly for tree discovery, but in terms of production systems and matching, we need the full (all sources) tree.
A multiplicity of source types is a mixed blessing.

- Customers do not know where the treemaker is getting his information. This is especially important when looking for, or finding, errors in the tree. Why is N-P189.2 in the wrong place on ISOGG's haplotree? Did some test go wrong? Or did someone make an assumption (perhaps based on an inadequate academic paper) without any test evidence?

- The treemaker is prone to throw away SNPs that are problematic in any one of the technologies. Thus, for example, FTDNA promptly threw away YP977 (in the R1a haplogroup) because it was unreliable in the SNP pack, even though YP977 is detected well enough in the Big Y and individual SNP tests. Thus, those who had paid good money for individual YP977 SNP tests were left in the lurch.

My point is what I said before: Each haplotree has its advantages and shortcomings.

miiser
04-29-2016, 08:30 PM
On the topic of Big Y and Y Elite being dead... Look at the cost delta of going with GenomeGuide versus Y Elite. This is about the cost of Family Finder. You lose a bit of Y coverage, but it's still more than most Big Ys. You gain the full mtDNA sequence and most likely every thing of note for the Family Finder. The autosomal results can then be formatted for upload into something like Gedmatch or pose as an Ancestry test into FTDNA.

That's why WGS is becoming interesting. The price isn't too much more than piecing your complete genome together with FTDNAs offerings. In some ways it's superior and others there is an inconvenience factor.

The other difficulty is the sales price. Selling an $895 product to newbies is more difficult than getting them to jump at a $99 product based on a 30 second TV ad or web banner. I almost think the WGS sellers need to start offering a loss leader or break even product to attract customers and be more competitive with the entry price - offer a $99 product to do a limited WGS test (2X or 4X coverage), even though the cost is greater than this. Then give the customer some subset of the data, such as terminal Y haplogroup, and offer more data with "upgrade" purchases. The customer would then be paying for data that the test company already has in storage and can provide without cost to themselves. Then they can lure the customer along to greater expenditure, nickel and dime like as FTDNA does, with possible eventual upgrades to 30X coverage.

TigerMW
04-29-2016, 09:07 PM
A multiplicity of source types is a mixed blessing.
It causes normalization and integration challenges but there is a an advantage to having as many sources as possible. There is some cross-checking validation as well as more completeness in the tree. That's what I think we want, a complete tree.


- Customers do not know where the treemaker is getting his information. This is especially important when looking for, or finding, errors in the tree. Why is N-P189.2 in the wrong place on ISOGG's haplotree? Did some test go wrong? Or did someone make an assumption (perhaps based on an inadequate academic paper) without any test evidence?
FTDNA says they will not add an SNP to their haplotree unless at least two people are derived for it in their database and it is found consistently according to the section of the tree it fits descendant to, parallel to, and ancestral too. Their proofs are much the same as ISOGG's. However, FTDNA should publish their proof documentation, at least on the GAP tool. I have the same complaint of ISOGG's tree, they need to publish their proof documentation. The truth is neither FTDNA's nor ISOGG's tree is suitable for references in academic papers as the proofs / raw data is not available from either.


- The treemaker is prone to throw away SNPs that are problematic in any one of the technologies. Thus, for example, FTDNA promptly threw away YP977 (in the R1a haplogroup) because it was unreliable in the SNP pack, even though YP977 is detected well enough in the Big Y and individual SNP tests. Thus, those who had paid good money for individual YP977 SNP tests were left in the lurch.
This is true. In some circles people say that Sanger Sequencing is the gold standard but the truth is a subclade is a subclade regardless of the SNPs and the technologies abilities to read it.

It is also true that sometimes an SNP is thrown out for better reasons than are explained, which is a communications problem. I've seen this happen with a major SNP in L21. At first I though FTDNA was crazy or their tests were defective but I've since found that there actually are back mutations in this case. I wish they would have explained that better to me in the first place.


My point is what I said before: Each haplotree has its advantages and shortcomings.
Of course, but I will add that a the fuller the tree, the better. Accuracy is also critical and so the trade-offs begin.

TigerMW
04-29-2016, 09:21 PM
...

Vince T followed Thomas post with ,"Yup, the laggards that have been putting off Big-Y and Y-Elite have chosen wisely. They need to wait until WGS 30x comes down in price, or even longer once nano-pore (whole-molecule) sequencing becomes viable, cheap, and ultimately succeeds NGS in quality"

The implication is FGC Y-Elite is dead. Do existing Y-Elite customers get an upgrade price with full credit for their investment or a data migration/conversion?

Of course Thomas and Vince are implying Big Y is dead too, but FTDNA may not agree. There are no indications that they do. They have their own plant and equipment so I don't know how we know their real costs and capabilities....

I should include the FGC's principal's comments from that thread.

Justin Loe wrote,
"We started working towards the goal of making a consumer whole genome test available in December 2011. That was the original goal, before FGC formed. I contacted a variety of organizations to get that going, and at the time it wasn't possible to sell a whole genome test for under $3,000. FGC was formed after I realized that we couldn't get this done as individual hobbyists, i.e. that companies wouldn't sell whole genome tests to us as individuals.

My goal always was to bring a whole genome test to market. We released a Y chromosome test because it was more practical at that time. Once the xTen was released we were able to start doing pilot sales of whole genome tests in 2014.

We set-up FGC partially because a certain other company and others weren't willing to sell whole genome tests, and I personally wanted that to be available. The Y chromosome test has been a stepping stone toward that goal.

Of course as Thomas says, the whole genome test contains all of the Y chromosome data of the Y Elite test. However, people still have a certain reluctance in going for a whole genome test, possibly because of the health information in it.

It is worth pointing out that for genealogy purposes, one could partition out the ancestry related markers, such as the Y chromosome, mtDNA, and 3.6 million SNPs found while keeping the rest of the data private. One could create an ancestry database built on a combination of comprehensive Y data, full mtDNA sequences, and 3.6 million SNPs."
https://groups.yahoo.com/neo/groups/R1b1c_U106-S21/conversations/messages/42653

It looks like he consider FGC Y Elite as a "stepping stone" to "whole genome tests." Y Elite may be a learning experience stepping stone for FGC and I hope he considers how he can make this a customer stepping stone so they carry their investments forward electronically.

I like that he is talking about databases too, "One could create an ancestry database". Customer value for genetic genealogy ends up being derived from databases that support comparisons between people.

Cofgene
04-30-2016, 01:47 PM
One of the unspoken arguments for WGS results over Y-elite or Big-Y surrounds the analysis of highly homologous regions. With higher quality WGS results advances can be made in "substractive" methodologies to open more regions on the Y for routine variant identification. Current homology restrictions are derived from specific technology limitations for only looking at a specific location.

TigerMW
05-02-2016, 03:45 PM
I'm trying to refine the why Big Y explanation and articulate data management, which I think is often overlooked or misunderstood. Please see item #5 below. I think that Y SNPs is an area that FTDNA has demonstrated data integration, albeit slowly. They've done the same with STRs as they add on more. I think the myFTDNA dashboard can be handy as an integrated view. For instance, since I've done Family Finder I've going now into my cousins' match list and looking for my surname, surname variants or geographical commonalities with my MDKA. Then I can contact those people to inquire about Y DNA testing in their family along with their genealogy. Actually, I'm behind on this. Others are contacting me.

........

We need more close matches more than anything. We need more people testing more deeply.

The biggest issues aren’t technical features of new tests. If we want genetic genealogy to truly expand the vendors should focus more on their total offerings, not just their bare-bones test features.

The essence of genetic genealogy is not to know more about your Y chromosome in isolation, but to compare the same locations or patterns across more people who are potentially related. Don't get me wrong, more locations covered is good, but what we really need is more testers in the matching databases!

These are the Big 8 advantages for Big Y. The culmination of these advantages provides what’s really important. Big Y provides the largest matching database of genetic genealogy results, not anonymous individuals or ancient skeletons. Big Y is the clear market share leader with several thousand Big Y results just for R1b. Its marketplace momentum has a life on it’s own no matter what we say here. Big Y has reached critical mass.

1. Lowest absolute entry price (attracting the most genetic genealogy testers).

2. Faster test processing time (probably because they do it in-house).

3. Convenience because of multiple test types (Big Y, STR panels, SNP Packs, mtDNA, Family Finder, etc.) available based on DNA samples that are safely stored.

4. A durable project management platform for sharing and categorizing information that is critical to the newbies, future generations and non-technicians.

5. Integrated and backward-compatible data management, which is the foundation of a large matching database. Old Deep Clade, Advanced Test SNPs, SNP Packs and Big Y results are all used for haplotree and haplogroup assignment and project display.

6. Early feeder of your SNPs into a broader set of offerings.

7. Excellent record and prospects as a proven on-going, self-funding concern.

8. The culmination – the largest accessible and compliant matching database for genealogy from the clear market share leader. This is a database of real people who’ve shown some inclination for spending their own money on DNA testing.

FGC Corp
05-02-2016, 04:31 PM
deleted by me.

TigerMW
05-02-2016, 04:46 PM
deleted by me.
Please don't take the opinions I share as something with deep and dark intent.

I AM very impressed with what I know about FGC products and I expect them to be vital to the industry and hopefully to science as well.

I just am trying to provide a point of view that is often lost among technicians. How do we get value out of this stuff? How can we improve genetic genealogy? It's just my opinion that low absolute pricing and lots of participants are a good things for those purposes. FTDNA is not a bad player in supporting genetic genealogy.

FGC, no doubt, has and will provide great things for people who need the kinds of services that FGC provides.

FGC Corp
05-02-2016, 04:50 PM
I have reached the end of my participation in this thread.

That is all.

JamesKane
05-02-2016, 05:57 PM
Does FTDNA really have the lowest cost to entry for NGS? The 4x WGS option is less expensive, will have similar results as the 1KG tests in ytree.net, and can be upgraded incrementally. The 2x is less expensive yet.

There are no absolutes or one best strategy. Big Y is good until you fall into a clade with a plateau because your branch is defined by SNPs outside it's target. Only testing men who already have an established ancestor around 1800 may continue to break them up. If you already know that on paper, it's more economical to test the 6-8 private SNPs directly. Of course that also requires another lab. Thus defeating all the purported benefits of FTDNA's database, since their staff is too busy chasing Packs.

The only reason I continue to get involved with these discussions is the continued claims of others that people should focus on FTDNA's products. The Big Y is all you need claims from speakers and admins on forums is simply incorrect. Multiple testing platforms that can be compared are crucial. Otherwise, you wind up dumping thousands down branches because the most popular one doesn't look in the correct locations. The team sport requires a few leaders to step up to the plate and score a few extra points to reveal the limitations quicker.

If no one is aware of the options, they can't exercise them.

TigerMW
05-02-2016, 06:26 PM
....
There are no absolutes or one best strategy....
Absolutely.


The team sport requires a few leaders to step up to the plate and score a few extra points to reveal the limitations quicker.
If no one is aware of the options, they can't exercise them.
Agreed. We need what I've been calling lead-explorers. Perhaps I should think of it as we have lead-scouts and surveyors but there is a premier group that are the lead-discoverers.

In any case, we need way more folks.

leonardo
05-02-2016, 08:19 PM
Does FTDNA really have the lowest cost to entry for NGS? The 4x WGS option is less expensive, will have similar results as the 1KG tests in ytree.net, and can be upgraded incrementally. The 2x is less expensive yet.

There are no absolutes or one best strategy. Big Y is good until you fall into a clade with a plateau because your branch is defined by SNPs outside it's target. Only testing men who already have an established ancestor around 1800 may continue to break them up. If you already know that on paper, it's more economical to test the 6-8 private SNPs directly. Of course that also requires another lab. Thus defeating all the purported benefits of FTDNA's database, since their staff is too busy chasing Packs.

The only reason I continue to get involved with these discussions is the continued claims of others that people should focus on FTDNA's products. The Big Y is all you need claims from speakers and admins on forums is simply incorrect. Multiple testing platforms that can be compared are crucial. Otherwise, you wind up dumping thousands down branches because the most popular one doesn't look in the correct locations. The team sport requires a few leaders to step up to the plate and score a few extra points to reveal the limitations quicker.

If no one is aware of the options, they can't exercise them.
Perhaps this is applicable to a few branches, such as some found in R1b? In R1a, I don't believe we have reached the point of saturation. I believe this to be true at least for my branch. There, we are still at the stage of getting as many as possible to test. The entry level price and having a sample already on hand are advantages of FTDNA, along with the administrators at the R1a project, who personally review people results and make suggestions that make the most sense - both for genealogically and economically (not everybody can afford these expensive products). I hope to get to the point someday where the BigY truly does not suffice and there is another product to further the cause.

jjtjr
05-02-2016, 09:51 PM
I am a DNA neophyte. I find this to be a very interesting and informative thread.

I had a 23 & Me autosomal test a couple of years ago. That lead me to a distant cousin who told me that I should test my Y-DNA. I had my first Y-DNA test about a year and a half ago.

I started doing YSEQ panels and individual SNP tests. I then did FTDNA STR tests up to 111 STR's and finally the Big-Y. I had my Big-Y results analyzed by both James Kane and YFull.

As James Kane mentioned above:


There are no absolutes or one best strategy. Big Y is good until you fall into a clade with a plateau because your branch is defined by SNPs outside it's target. Only testing men who already have an established ancestor around 1800 may continue to break them up. If you already know that on paper, it's more economical to test the 6-8 private SNPs directly. Of course that also requires another lab. Thus defeating all the purported benefits of FTDNA's database, since their staff is too busy chasing Packs.

I recently tested 5 cousins for my current terminal SNP at YSEQ. YSEQ has also developed a panel based on my Unique Mutations found in the Big Tree, James Kane's matrix and YFull. I am testing one of my cousins for that.

If I have matches there, I will probably test a cousin for the Big-Y mainly for their data bases and projects.

After reading this thread I might do a FGC YDNA test on myself in the future, since they are the most complete that I am aware of.

I don't care about the scientific value of DNA. I have traced my Family on paper back until about 1800. I am doing DNA testing to learn more about them in the 1600-1700's. That might take time, but I intend to use any tools that are available to me.

This thread has helped me understand some of those tools.

RobertCasey
05-02-2016, 11:20 PM
I recently tested 5 cousins for my current terminal SNP at YSEQ. YSEQ has also developed a panel based on my Unique Mutations found in the Big Tree, James Kane's matrix and YFull. I am testing one of my cousins for that.

If I have matches there, I will probably test a cousin for the Big-Y mainly for their data bases and projects.

After reading this thread I might do a FGC YDNA test on myself in the future, since they are the most complete that I am aware of.

I don't care about the scientific value of DNA. I have traced my Family on paper back until about 1800. I am doing DNA testing to learn more about them in the 1600-1700's. That might take time, but I intend to use any tools that are available to me.

This thread has helped me understand some of those tools.

Good to see another genealogist on this thread. My main reason for testing is to break through a brick wall in the late 1700s where we can not tie around 40 known lines to together via traditional genealogy, so your quest is very similar to mine.

I need some more information about your line. What part of the haplotree do you belong to ? This will make a big difference on any testing strategy. Also, how many 67 marker submissions belong to your cluster and are they all proven relatives or only connected by surname and geography.

For my Casey cluster under L226 (which is around 1,400 years old), I have discovered five new branches with my Full Genomes Y Elite 1.0 test:

L226>FGC5660>FGC5628>FGC5659>FGC5647>FGC5639

The first three are major older branches under L226 that are shared by a lot of surnames. The last two Casey only mutations - but both are at the very top of the cluster. FGC5647+ finally confirms that a Kersey submission shares this mutation in common with all the Casey below but not the Carey submissions above. FGC5639 is Casey only and Kersey tested negative. But these YSNPs are right at surname creation but have not split up our cluster to date. Oh so close. So were getting really close and still have a lot of testing to see if more branches are discovered.

I have run into several issues though. It is very hard to convince others to test (we have around 25 67 marker submissions in our cluster). I also have 12 private YSNPs that have not been tested by 48 other Big Y tests under L226. So these private YSNPs need to be tested after the original seven private YSNPs tested by Big Y are thoroughly tested. I have drafted another cluster member to take a second FGC Y Elite 2.1 test which will create even more private YSNPs to test due to being more coverage than 1.0 and being in a different part of the cluster.

But I still will have a problem with all of these higher coverage YSNPs. My original 1.0 test had an amazing 50 % that were equivalents of L226 which is around 1,400 years old. It is anticipated that these higher coverage private YSNPs will also be yet 50 % more L226 equivalents if the ratio holds up. I have two possible approaches where I think one is much better. I have been testing a fairly close matching Carey but he has tested negative on five private YSNPs to date. I could use him as control test for L226 equivalents - if he tests positive, they must be L226 equivalents and if he tests negative, they would be new private YSNPs test for our cluster. The second method is to test somebody who is part of a very early branch of L226. There are some (less than 5 % of L226) that fall into this category. I could use this person as control test as well to filter out the 50 % that will L226 equivalents.

I think testing the older branch is much better since there should be two major new branches discovered between L226 and my FGC5647. Using a closer match would not reveal these possible major branches as they would be lumped in with the L226 equivalents and would require more testing to separate the two. Testing the older branch, I would be left with major branches and private YSNPs testing negative - a much better outcome.

So I really need to test the 12 private YSNPs not tested by Big Y at YSEQ (who tests these unlike FTDNA). This would cost $210 to filter out the L226 equivalents. A variation of this option would be sponsor the upgrade cost for replacing another Big Y test with a third Y Elite - the delta is $200, so we both get a lot more. I found a person who was interested but he is not real responsive to date since he was still considering NGS in general. The original person who NGS tested and created the oldest branch - refused to test my private YSNPs at my expense.

So it is real important to know your haplogroup and any know bottlenecks for your predictable single signature haplogroup. It is pretty common for a lot of these private YSNPs as being equivalents of some older YSNP. You really need to understand what testing is done well above your cluster to develop a proper strategy. You may not have my bottleneck issue of L226. But you need to know so you do not waste a lot of time and money testing equivalents and thinking it is a positive private YSNP.

MitchellSince1893
05-02-2016, 11:50 PM
For genealogical purposes I think you will be better off doing Y Elite vs BigY as you will want to find every novel SNP you can. Missing just one SNP can mean missing 5 generations on average (30 years per generation x 5 generations =150 years...roughly the average time per SNP mutation). Which would be very important if the SNP in question was the most recent.

That was a big reason why I did FGC Y Elite after doing BigY.

jjtjr
05-03-2016, 12:38 AM
Robert,

I am R1b>L21>DF13>Z255>Z16437>A557>A558>Z29008.

I currently share Z29008 with one other person on the Big Tree. Two of my cousins are also positive for Z29008 from the YSEQ single SNP test and the other three are being tested now. One of the positive cousins is being tested for the YSEQ panel right now.

I am lucky enough to have a few smart people counseling me on DNA. They have told me that tests like the Big-Y or FGC Y-Elite would yield much more information in the future than individual tests. I am sure that they are correct. Neither the Big-Y or Y-Elite are inexpensive.

My thought is that since I already know these 5 cousins (1-1st cousin 1x, 1 - 2nd cousin 1x, 3 - 3rd cousins) are related to me that I would save money and start by testing for my current terminal SNP of Z29008.

I assume that if I get matches in the YSEQ panel downstream of Z29008 that would it could lead to more information. I know that the problem with doing it this way is that information is not disseminated to others. That is why I will probably test 1 cousin for the Big-Y.


It is very hard to convince others to test

You are right there! It was hard to convince any these of these cousins to take a DNA test and I am paying for the tests. That is why that I would choose the Big-Y over the FCG Y-Elite, the cost, the database and the FTDNA Projects.

I am sure that the best way to go would to test all 5 cousins for the Y-Elite, but I can't afford to do that. Maybe one or more of my cousins will get interested in DNA and go further on their own.

I know that a year ago I thought that I would never pay for the Big-Y for myself, let alone someone else. Hopefully in the future I will take a more complete test.

RobertCasey
05-03-2016, 04:04 AM
I am R1b>L21>DF13>Z255>Z16437>A557>A558>Z29008.

My thought is that since I already know these 5 cousins (1-1st cousin 1x, 1 - 2nd cousin 1x, 3 - 3rd cousins) are related to me that I would save money and start by testing for my current terminal SNP of Z29008.


Testing such close cousins with YDNA may reveal very little information since you are so closely related. It is always best to test multiple sons of your oldest proven ancestor for this line to gain the most insight. I would switch your YDNA testing to remote cousins where YSTRs and YSNPs have more time to mutate.

I see that Alex shows Miller 219493 and Treacy B348804 as the only two testing positive for Z29008. They have a pretty good signature 439 >= 13, 389-1 >= 13, 458 <= 17, 449 >= 30, 557 >= 17 and 446 >= 14 at 67 markers. However, they have a very large genetic distance of ten. Neither of the above include CDY markers that mutate too fast to be included. The genetic distance is a pretty big red flag indicating that Z29008 is probably pre-surname creation in time frame (over 1,000 years old). Both have different surnames as well and their father, A557, have Creamer and McCarthy as surnames further suggesting these branches are probably older than 1,000 years.

Your really need build your cluster with people that are remotely related and preferably no known genealogical connection established. Also, if you have researched this line, you could test a good possible genealogical match and confirm the connection with YSTRs. Have you tested to 67 markers very recently as I did not find any other submissions that are 67 markers in my spreadsheet or Mike's spreadsheet ? What is you FTDNA ID ?

YSNPs only mutate around every three generations, so all your cousins would at best match one private YSNP that you have NGS tested. Is your oldest proven ancestor's surname Miller or Treacy ? I think testing your cousins for Z29008 is not that necessary as they will all probably test the same as you when a different person tested the NGS test.



9387058-G-T




+


7109717-A-C




+


22042869-C-T




+


26707300-A-AT


P1_g2

+


21562790-CA-C




+


9172337-C-T




+


8890393-G-A




+


8468136-T-C




+


8546926-G-C






If you drill down on Alex's chart, it looks like there are other private YSNPs that have not been assigned any label to date. This information makes the YSTR and YSNP information more consistent. You could label these private YSNPs and submit Wish a YSNP at YSEQ and possibly find both positive and negative results creating a new branch. If you test positive for most of these YSNPs, this mean a closer relationship. If you test negative for most, then are more distantly related. You also need to contact Miller and Treacy for more guidance.

jjtjr
05-03-2016, 12:42 PM
Robert,

My name is Jim Treacy, my FTDNA # is B348804. The cousins that I have tested so far at YSEQ are descended from 4 different sons of my MRCA, my Gr.-Gr.-Grandfather born circa 1794. I only know of 2 possible 4th cousins. I am trying to contact them.

I believe that the Time to Most Recent Common Ancestor between D. Miller and myself is 750 years. It is about the same for Creamer & McCarthy with who I share A557.

As I mentioned I am fairly new to DNA testing. I think that I have a basic knowledge of SNP's, but I know less about STR's and their significance. I am lucky to have a few people helping me learn more about DNA.

I completed Y-67 testing in the beginning of Feb. 2016 and Y-111 testing in the end of Feb. 2016. I am only show at 37 markers in Group E of the McCarthy Project. I did not know that I am shown on your spreadsheet.

I have been told that testing my known cousins for Z29008 might be a waste of money since barring a NPE, that they are probably positive for Z29008. I am doing it because it is inexpensive and might form a small cluster of Treacy men positive for that SNP. My thought is that if I get matches in the panel including my unique SNP's that YSEQ has developed downstream of Z29008, that I will then test a cousin further.

In addition to the cousin being tested for the YSEQ panel, I also have his son being tested for Y-37 at FTDNA. This should let him join various FTDNA Projects as well as upgrade in the future. If his father is positive in the YSEQ panel for SNP's downstream of Z29008 I will probably test him for the Big-Y as I mentioned.

I know that there are other Treacy's living in the East Galway area where my Family lived. I don't know of any connection with them. Am I correct that you think that I should contact some of them and ask if they would take a DNA test? Would just testing for the Z29008 SNP be enough to show a connection?

I am still learning about DNA testing and that is one of the reasons that I appreciate this thread. It helps me learn the value of using different DNA tests.

Thanks for giving me more information.

RobertCasey
05-03-2016, 02:08 PM
Robert,

My name is Jim Treacy, my FTDNA # is B348804. The cousins that I have tested so far at YSEQ are descended from 4 different sons of my MRCA, my Gr.-Gr.-Grandfather born circa 1794. I only know of 2 possible 4th cousins. I am trying to contact them.

I believe that the Time to Most Recent Common Ancestor between D. Miller and myself is 750 years. It is about the same for Creamer & McCarthy with who I share A557.

I completed Y-67 testing in the beginning of Feb. 2016 and Y-111 testing in the end of Feb. 2016. I am only show at 37 markers in Group E of the McCarthy Project. I did not know that I am shown on your spreadsheet.

I have been told that testing my known cousins for Z29008 might be a waste of money since barring a NPE, that they are probably positive for Z29008. I am doing it because it is inexpensive and might form a small cluster of Treacy men positive for that SNP. My thought is that if I get matches in the panel including my unique SNP's that YSEQ has developed downstream of Z29008, that I will then test a cousin further.

In addition to the cousin being tested for the YSEQ panel, I also have his son being tested for Y-37 at FTDNA. This should let him join various FTDNA Projects as well as upgrade in the future. If his father is positive in the YSEQ panel for SNP's downstream of Z29008 I will probably test him for the Big-Y as I mentioned.

I know that there are other Treacy's living in the East Galway area where my Family lived. I don't know of any connection with them. Am I correct that you think that I should contact some of them and ask if they would take a DNA test? Would just testing for the Z29008 SNP be enough to show a connection?

For YSTR testing, I would only encourage children and grandchildren of you oldest proven ancestor. But if your closer cousin does test, he needs to be 67 markers but not really 111 markers until your cluster is much bigger (10 to 20 testers). Your off modal mutations from the L21 modal is your most important criteria for relatedness - then genetic distance is sanity check that convergence is not present. Your signature for Z29008 is currently 439 >= 13, 389-1 >= 13, 458 <= 17, 449 >= 30, 557 >= 17 and 446 >= 14 at 67 markers. Since this is based on only two submissions, it will probably get a little smaller over time.

My main concern is that you (B348804) and the Miller (219493) have a genetic distance of 10 which indicates the relationship is quite old - way before any genealogical research could help and probably earlier than surname creation (1,000 AD for most Irish). Also, Miller is a English name and Treacy is an Irish name which is another red flag.

TMRCA's are not very accurate for under 1,000 years. Once you get into this time frame, some common sense logic about NPE rates is more reliable - adjusting for any bias in testing. Your known Treacy line would only count once and the Miller line would count. So their currently would have to be a 50 % NPE rate for the last 1,000 years. Once your cluster gets a little bigger, comparing surnames lets you know the age of Z29008 is relative to surname creation - around 1,000 years for most Irish and Scottish names (does vary though) and 200 or 300 years more recent for English surnames.

Testing closer cousins for Z29008 will be unlikely to reveal much but as you say testing only Z29008 is $17.50 at YSEQ. Again, building your cluster under Z29008 is the most important task as you really need more than two YSTR tests. Once you find a son or grandson's line of your oldest proven to test for 67 markers first. You could then test your private YSNPs at YSEQ and will probably find another branch below Z28008. If you do not assign any label to the mutations, YSEQ will probably add Axxxxx labels for you. They will create new YSNP tests for each of your private YSNPs via Wish a SNP at $1 each.

If you find Treacy's in smaller Irish towns where your ancestors lived, sponsoring the 67 marker test would be an excellent way to build your cluster. Galway is getting to be a larger town but if their roots in the other two smaller towns or other small towns north and northeast of Galway, that would raise the odds of being related. You should also join the Treacy surname project (probably lumped into Tracey). If you find YSTR matches with the L21 off modal values, getting them to upgrade to 67 and test your Z29008 would be another way to expand your cluster.

jjtjr
05-03-2016, 03:31 PM
Robert,

Thanks for the information on my Z29008 signature. Hopefully more people will test positive for it in the future. I will upgrade my cousin's STR test to Y-67 when FTDNA has another sale.

I believe that D. Miller's Gr.-Grandfather changed his name. That might explain an English versus Irish Surname.

At 67 markers I had a genetic distance of 6 or 7 with five members of the Gleason/Gleeson clan. At 111 markers I had no matches. It was explained to me that this was convergence or back mutation in my DNA.

I originally submitted 28 unique SNP's to YSEQ. After analyzing them they came up with a panel of 14 SNP's. They have added Axxxxx labels to many of them.

I am a member of the Tracy DNA Project and others. My 3rd cousin will join the Tracy Project when he gets his Y-37 results.

My Family is from the Tynagh area near Portumna in East Galway. It is a smaller town. I will see if I can contact Treacy's from that area who I don't know that I am related to. For now at least, if I do find a Treacy willing to be tested from the area I will just test for the Z29008 SNP, mainly because of cost.

TigerMW
10-15-2016, 09:54 PM
I have refined this a bit and normally wouldn't go back to an old thread like this but I see Paul G referring to this thread on Facebook currently so I'll add in my two cents.

This is now in the FAQ section of the R1b project web pages. For people that are true researchers, academics or have a very specific purpose when they get a DNA test, they have to look at speeds and feeds of the test, perhaps foremost.

For average folks interested genetic genealogy, a big deal about Big Y is that it is integrated into a genetic genealogy platform. There are no charges or annual subscription rates and you get the following.

https://www.familytreedna.com/groups/r-1b/faq#/FFTDNA

"Why DNA testing at a full platform genetic genealogy company?

FTDNA provides much more than bare bones DNA tests. They have a true laboratory and our not just a test remarketeer. They have no annual subscription charges, but provide a comprehensive platform for genetic genealogy including:

1. A robust project management web based platform for sharing and categorizing information that is critical to the newbies, future generations and non-technicians.

2. The largest accessible and compliant matching database for Y genetic genealogy. This is a database of real people who’ve shown some inclination for spending their own money on DNA testing.

3. Integrated and backward-compatible data management, which is the foundation of a large matching database. Old Deep Clade, Advanced Test SNPs, SNP Packs and Big Y results are all used for haplotree and haplogroup assignment and project display.

4. Convenience because of multiple test types (Big Y, STR panels, SNP Packs, mtDNA, Family Finder, etc.) available based on DNA samples that are safely stored.

4. Family tree and GEDCOM support integrated to DNA matching results.

5. Reasonable pricing which helps attract the most genetic genealogy testers.

6. Faster test processing time for SNP Pack and Big Y tests because they own their lab and test in-house.

7. They are an early feeder of your SNPs into a broader set of offerings including products from National Geographic Society organization sponsored tests that are used all over the world.

8. Excellent record and prospects as a proven on-going, self-funding concern."

TigerMW
10-15-2016, 10:28 PM
I guess I was right in saying they were too conservative, at least in your opinion. As I noted earlier, "It will be a moot point soon anyway as you'll have downstream SNPs marking your new terminal haplogroup."

You made a wise move having both you and your compatriot get Big Y. You've found a new line of SNPs that form a new branch, no doubt about it....

Just curious. I would expect your haplogroup label to be R-BY9002 because I see BY9002 on the FTDNA haplotree. Has it happened? If not we may need to email a fellow named Michael Sager to force the haplogroup update.

I looked at the Big Tree and see you have phylogenetic equivalent SNPs in this block. I'll see what I can do about that.

JamesKane
10-15-2016, 11:52 PM
2. The largest accessible and compliant matching database for Y genetic genealogy. This is a database of real people who’ve shown some inclination for spending their own money on DNA testing.

This may be true but it overlooks a very important aspect. The FTDNA database is limited to their tests. Branches continue to be defined by data sets such as 1000 Genomes, PGP: Harvard, the multitude of contemporary academic sources, and just as importantly their competition. Yes, there are thousands of the Big Ys but there are also thousands of samples out there they do not acknowledge.

The Y DNA tree is being built out by volunteers, scientists, and even a for profit company which is not FTDNA. These are the folks Big Y testers turn to make sense of the data dump supplied by FTDNA and it's intractable Big Y matching. Given the collaborative nature of the Internet having a centralized testing vendor is not necessary nor is it necessarily even desirable.

Let's also not forget the demographic traditionally associated with genetic genealogy. This is an aging population by in large. How many of those people who have tested before are still among us or passed on the necessary permissions to continue testing on their samples?

RobertCasey
10-16-2016, 04:25 AM
"Why DNA testing with several genetic genealogy companies?"

FTDNA is by far the largest and the most complete testing company for the genetic genealogical community. But with its larger size, also comes higher testing costs and less innovation. They also have the most comprehensive database but the quality of that database has numerous problems that have not been addressed. The main bright spot is their recent focus on their haplotree.

1. Database - There are three main components of YDNA FTDNA databases: a) YSTR reports; b) YSNPs reports; c) their YSNP haplotree. Their YSTR reports are reasonably complete to work with except their roll-out of increased privacy settings continue to default to "private" and allowing projects to greatly reduce accessibility of data which drive up our overall costs for our community and lowers their operating costs by trimming down the database size that is accessible. Their YSNPs reports do not include numerous relevant YSNPs found in Big Y tests. None of the private YSNPs are posted, many of the downstream relevant YSNPs are not uploaded and they continue to bloat the database with many upstream YSNPs that are not that relevant. They also ignore many sound YSNP branches found in the ISOGG haplotree as well. Their haplotree has been transformed from being very poor to leading edge in the last few months - this is a hard nut for all of us to keep up with. However, they no longer offer individual YSNP testing of private YSNPs which greatly decreases the robustness of their YSNP reports. YSEQ does have a reasonable database for YSNPs - it is just another database to contend with. Since FTDNA does not offer testing private YSNPs, the YSEQ database is superior in content by having more relevant private and more ISOGG YSNPs. Full Genomes has no database for NGS results but FTDNA uploads only minimal YSNPs from their Big Y tests.

2. Overall testing costs - Full Genomes offers 30 % more coverage for 30 % more cost of the Big Y. The price per YSNP discovered is the same and Full Genomes has made five or six significant upgrades while the larger company, FTDNA, has kept the scope of their test the same. This issue is very similar to comparing the 67 marker test to the 111 marker test. Resolution does matter. YSEQ testing of individual YSNPs is 50 % the cost of FTDNA. YSEQ first introduced YSNP panels which was a huge step in reducing YSNP testing. With robust testing of private YSNPs, discovering branches via YSEQ is running at 10 % of the cost of your approach of only Big Y tests for the L226 project to date. This lower cost approach, makes branch discovery in the reach of more of the genetic community that can not afford the Big Y test. We have over 350 private YSNPs under L226 - that is a huge opportunity to discover new branches. However, FTDNA just made a quantum leap in the L226 and L555 SNP packs by adding a lot of private YSNPs - but at the same time, they polluted their haplotree database (and YSNP reports) by adding all these private YSNPs as branches which is causing great confusion.

3. Innovation - FTDNA has lost its innovative edge in most of the important issues: 1) 23andme introduced atDNA testing and then we get Family Finder. 2) Full Genomes introduced the first NGS test 15X coverage of Walk the Y - then we get Big Y; 2) FTDNA decided to drop their support of YSNP testing even though YSEQ YSNP testing is revealing many new branches without the need for expensive Big Y tests; 3) YSEQ introduced SNP panels and then FTDNA announced SNP packs. However, with the new Mass Array YSNP pack tests, the significant rollout and updating of SNP packs and now inclusion of private YSNPs - FTNDA now has a superior offering again. Within the next 12 to 24 months, Full Genomes will rollout a Whole Genome Sequence test with longer read lengths that will be at the price of the current Big Y. This test will include 35 % more coverage than the Big Y test, include all 111 YSTR markers plus 300 more YSTR markers and will include all atDNA data as well. The FTDNA haplotree remained so back level for such a long time, it was rarely used and ISOGG reigned. But ISOGG has become so restrictive and requires so much documentation, its haplotree has lost its impact. Alex Williamson came out with his Big Tree (R-P312 only) and then FTDNA later rolled out their improved haplotree.

4. Overall lower costs - Testing of private YSNPs and robust SNP pack testing discovers new branches at 10 % of the cost of the proposed methodology of FTDNA only Big Y testing. The FTDNA concentric approach ignores the 1,000s of private YSNPs that can be tested directly for $75 to $150 vs. $575. The Big Y is always superior quality test as it discovers yet more private YSNPs - but it costs 10X to discover branches. In the next iteration or two of Full Genomes WGS tests, it will be a bundled test that includes NGS testing (35 % more), YSTR testing (300 % more including all 111 markers) and atDNA testing (1000 % more if it helps) - at the same cost of the Big Y. Full Genomes is already selling WGS testing for under $1,000 with 108 of 111 YSTRs at the current read length. They just recently announced 1,000,000 read length test which will reveal even more of the Y chromosome - not sure how much additional YSNP discovery this will provide but it is very innovative and will surely discover some major branches.

5. In house testing vs. outsourced testing - By using other vendors for testing, Full Genomes can afford to constantly use the best technology as it becomes available which leads to more innovation. With FTDNA having in house technology, they must amortize their equipment costs and then will be tempted to keep using the equipment longer to obtain more profits until competition forces their hand. In house does allow faster turnaround times and the potential to reduce costs after the equipment is fully amortized - but we have not seen any price reductions to date with Big Y.

Cofgene
10-16-2016, 11:03 AM
The value statement is incomplete and tilted in FTDNA's favor. FTDNA does not accurately report or utilize INDELs or MNP's in their output. That is where we have to tell them what they are missing. For Big-Y they scrub the data so hard one will get missing regions when the BigY is converted to the current human reference build. That is negative value! Those issues are not present in FGC results. Add in the additional results of getting mtDNA data and more STR calls one then sees that the value proposition moves solidly towards FGC. BigY results are missing important clade defining levels that are identified in WGS or FGC Elite results. Selected individuals with Big-Y results will need to do additional testing to properly position their result on the tree. Once again the need for follow-up testing even with a Big-Y result represents negative value. Individuals need to stop investing in the Chevy (Big-Y) and go with the WGS/Elite (BMW) solution as the best long term value for getting longterm returns.

lgmayka
10-16-2016, 11:18 AM
4. Overall lower costs - Testing of private YSNPs and robust SNP pack testing discovers new branches at 10 % of the cost of the proposed methodology of FTDNA only Big Y testing.
This may be true of your L226 project, but yours is a rare case. As YFull says (https://yfull.com/tree/R-L226/), the L226 clade diverged from its siblings 4300 years ago, but began to expand only 1400 years ago. Since then, the clade has grown extremely rapidly, first across Ireland and then across the globe. Thus, the R-L226 clade has two amazing advantages:

- Its 2900-year gestation period ensures that it is recognizable from Y-STRs alone. In almost any other clade or haplogroup, considerable Y-SNP testing is required to reach the specificity of a 1400-year TMRCA--at least one SNP pack, and often two.

- Its 1400 years of phenomenal population growth almost ensure that testing of new recent Y-SNPs will identify new branch points. But other clades and haplogroups which have not enjoyed such rapid recent growth do not yield such a plethora of new branch points. Consider, for example, R-Y2915 (https://yfull.com/tree/R-Y2915/), with a slightly older TMRCA of 1950 years. In this clade, 14 full-Y tests have found only 1 subclade. Testing of these men's newly discovered SNPs would have been an expensive waste of time, because almost all of them are unshared.

RobertCasey
10-16-2016, 03:23 PM
This may be true of your L226 project, but yours is a rare case. As YFull says (https://yfull.com/tree/R-L226/), the L226 clade diverged from its siblings 4300 years ago, but began to expand only 1400 years ago. Since then, the clade has grown extremely rapidly, first across Ireland and then across the globe. Thus, the R-L226 clade has two amazing advantages:

- Its 2900-year gestation period ensures that it is recognizable from Y-STRs alone. In almost any other clade or haplogroup, considerable Y-SNP testing is required to reach the specificity of a 1400-year TMRCA--at least one SNP pack, and often two.

- Its 1400 years of phenomenal population growth almost ensure that testing of new recent Y-SNPs will identify new branch points. But other clades and haplogroups which have not enjoyed such rapid recent growth do not yield such a plethora of new branch points. Consider, for example, R-Y2915 (https://yfull.com/tree/R-Y2915/), with a slightly older TMRCA of 1950 years. In this clade, 14 full-Y tests have found only 1 subclade. Testing of these men's newly discovered SNPs would have been an expensive waste of time, because almost all of them are unshared.

I agree that L226 has the ideal conditions for genetic genealogy but it is not as rare as you suggest under L21. M222 and L1335/L1605 are much larger with 45 equivalents each. L555 has 38 equivalents but was much less prolific - most are Irvine related. On the other end L193 (aka Mike Walsh) has a mere nine equivalents yet is another example of a robust haplotree beneath it. Of course Mike's help is another advantage and good leadership of haplogroups helps a lot as well. L226 is also blessed with three active researchers. The extra time to develop a more isolated signature is a factor - but there are a lot of L21 branches that are much younger with very strong signatures as well. So having a strong and isolated signature is the real advantage - more than just the bottleneck of time. Also, being prolific is another major factor - but L226 is medium sized compared to M222 and L1335.

There are many people hanging out there with few surviving male descendants. First congratulations for making it as 90 % of the male lines have probably died out completely in the 2,000 years. The massive invasion of R1b into Europe long after the last ice age just shows you how much technology plays a huge role in preserving you bloodlines. Also, being unique equates to less debate about relatedness. Under L226, we have a newly discovered issue which is a major bottleneck. Since the the growth of L226 only started around 1,500 years ago, there has not be enough time for YSTR divergence. These poor unlucky ones are similar to what you are addressing. These people has genetic distances of 1 to 3 at 67 markers but clearly belong to pre-surname clusters over 1,000 years ago. Around 20 or 30 % of L226 may fall into this "lack of divergence" where only YSNP testing will reveal relatedness.

TigerMW
10-16-2016, 03:48 PM
So user beware when
looking at terminal YSNPs in their YSTR and YSNP reports. There
are also terminal YSNPs routinely do not match their haplotree -
but the haplotree is so vastly improved, I welcome a few glitches
here and there.
I want to clarify how I count distinct branches. I do not count unique haplogroup labels. That would represent a duplication of branch counting where phylogenetic equivalent SNPs are involved.

I count actual branches on the haplotree. I copy/paste that into a spreadsheet and do a calculation or two and a sort to filter to just one row per one branch. In other words phylogenetic equivalent SNPs are NOT counted. Only the first or "lead with" SNP for a haplogroup (branch) is counted.

As of Friday, I counted 2,715 distinct branches on the R1b tree.

I have not tried to track the phylogenetic equivalent SNPs on the entire FTDNA R1b haplotree, except for the L21 branches. The average number of SNPs per branch (row) on the FTDNA L21 tree is 6.3.

If you multiple 2,715 by 6.3 you end up with over 17,000 SNPs on the haplotree. I don't think that is really fair as not all sections of R1b have submitted phylogenetic equivalent SNPs to addition to the tree. Still it gives you some idea of what the tree should look like.

In terms of project displays, this short haplogroup label issue has now come out to full expression. I would like to see FTDNA add back the long haplogroup labels as an option. I would like to see (and have recommended) that they provide a couple of different configurable subgrouping views. This way we could subgroup by haplogroup, by geography and also by testing recommendations. I recommended this over a year ago but I'm not holding my breath.... may be some day. LOL

The long haplogroup labels are actually there under the covers somewhere already. That's how they sort out their tree branches.

MitchellSince1893
10-16-2016, 04:04 PM
Mikewww,

Could you give me a response to my email or pm? Thanks

RobertCasey
10-16-2016, 08:19 PM
I have not tried to track the phylogenetic equivalent SNPs on the entire FTDNA R1b haplotree, except for the L21 branches. The average number of SNPs per branch (row) on the FTDNA L21 tree is 6.3.

If you multiple 2,715 by 6.3 you end up with over 17,000 SNPs on the haplotree. I don't think that is really fair as not all sections of R1b have submitted phylogenetic equivalent SNPs to addition to the tree. Still it gives you some idea of what the tree should look like.


I agree with almost everything in your post but there is one additional factor that I would add. The numbers of equivalents are very high (I have not confirmd 6.3 but that is reasonable). However, there is a huge variation between YSNPs that have genetic bottlenecks (with large numbers of equivalent YSNPs) and the more recent YSNPs. From Alex's chart (which does track fairly well), M222 and L1335/L1065 have 45 equivalent YSNPs, L555 has 38, L720 has 30, L226 has 25 and L193 has 9. If you filter out all the largest 50 branches, the average would drop to two or three equivalents per branch (with lots of variation within this grouping as well).

I also agree that there are lot of missing equivalent YSNPs that FTDNA and admins submitting changes just have not had enough time to add these equivalents. However, I do not think the value of equivalent YSNPs warrants inclusion priority over private YSNPs in SNP packs. We already have discovered four branches with the 50 private YSNPs in the L226 SNP pack and no branches from 50 equivalents with the first 25 Mass Array L226 SNP test results. We have already had 50 NGS tests where these equivalent YSNPs have remained YSNP equivalents. I think that we would have eight branches if 100 private YSNPs were included instead. NGS tests is really a preferred way to convert equivalent branches into haplotree branches.

If you only include the top 10 % of highest numbers of YSNPs with equivalents, your number would probably double or triple as the point I think that you were making is that there are a lot of older YSNPs where there huge numbers of equivalent YSNPs. For these older YSNPs, there are not enough private YSNPs remaining and equivalents would have a much higher odds of discovering a new branch. I really do not want to repeat the L459 and Z260 funds that could have been directed to more Walk the Ys. Equivalents do yield some new branches, but we should favor private YSNPs over equivalent YSNPs for younger YSNPs. For SNP packs that cover older YSNP branches, a liberal usage equivalents make more sense.

TigerMW
10-16-2016, 08:55 PM
Let's take the fixed/targetted SNP testing discussion over to the R1b SNP Packs thread in the R1b-general category. I'm not sure if everyone cares about R1b and this thread title gets beat to death. I think some of this on targetted SNP testing depends and the focus needs to change over time. This is very good topic that I've run into during the discussions on v2/refreshes for SNP Packs.

J Man
04-05-2017, 08:44 PM
I have had the Y Elite test done with Full genomes...At FTDNA on 111 markers I have an individual who mismatches with me on 19 markers. I do not think he will ever test with Full Genomes Y Elite but he probably would test with BigY since his sample is already at FTDNA...Would there be much of a point in having him tested with BigY to compare him to my Full Genomes Y Elite results? Would there be a lot of good SNP overlaps?

MacUalraig
04-05-2017, 09:12 PM
I have had the Y Elite test done with Full genomes...At FTDNA on 111 markers I have an individual who mismatches with me on 19 markers. I do not think he will ever test with Full Genomes Y Elite but he probably would test with BigY since his sample is already at FTDNA...Would there be much of a point in having him tested with BigY to compare him to my Full Genomes Y Elite results? Would there be a lot of good SNP overlaps?

It varies a lot from one individual to another but the coverage regions of BigY are documented so if you really want you can manually check the positions of all your FGC SNPs and get a good idea of whether they will be covered by a BigY.