PDA

View Full Version : YDNA predictor



Clarke
06-02-2018, 10:54 AM
Hi All

I have put my Ancestry raw DNA through Chris's YDNA predictor, these are the results. Is there any more detail apart from likely M222 that can be gained from this. I have plans to do a paid YDNA test in the future, but at present funds are short. Thanks in advance Clarke

2362423625

nuplix
06-02-2018, 11:38 AM
Try uploading your data on Wegene. It also predicts Y DNA and mt DNA.

Nqp15hhu
06-02-2018, 11:47 AM
Yeah, i've done the same thing and got 'M529' as a Haplogroup. Not really sure what to do with that information!

nuplix
06-02-2018, 11:49 AM
Yeah, i've done the same thing and got 'M529' as a Haplogroup. Not really sure what to do with that information!

Maybe, this helps you. Not sure!

https://yhrd.org/tools/branch/R1b-M529

ArmandoR1b
06-02-2018, 12:38 PM
Hi All

I have put my Ancestry raw DNA through Chris's YDNA predictor, these are the results. Is there any more detail apart from likely M222 that can be gained from this. I have plans to do a paid YDNA test in the future, but at present funds are short. Thanks in advance Clarke



There aren't any SNPs downstream from R-M222 that are tested by AncestryDNA. You've hit the end of the road until you get the planned testing.

ArmandoR1b
06-02-2018, 12:44 PM
Yeah, i've done the same thing and got 'M529' as a Haplogroup. Not really sure what to do with that information!

You should get the R1b-L21 Superclade Orientation Panel (https://www.yseq.net/product_info.php?cPath=27&products_id=4321) in order to find out which subclade you belong to. AncestryDNA doesn't test for DF13 or DF63.

RobertCasey
06-02-2018, 02:12 PM
You should get the R1b-L21 Superclade Orientation Panel (https://www.yseq.net/product_info.php?cPath=27&products_id=4321) in order to find out which subclade you belong to. AncestryDNA doesn't test for DF13 or DF63.
You can still find 50 % of L21 haplogroups via my very dated L21 SNP predictor tool:

http://www.rcasey.net/DNA/R-L21_SNP_Predictor_Intro.html

Also, Nevgen and my tool combined reliably predicts around 60 to 70 %:

http://www.nevgen.org/

So 60 to 70 % of the time, these gateway SNP packs are not necessary. If somebody wants to get my tool current, we could easily get prediction up to 80 % coverage of R-L21. My predictor tool stopped when Big Y was introduced, so there have been a truckload of new predictable YSNPs that could be added. If you have the time, I would teach you how to continually add branches and I would provide the web site to host it. I also now use two parameters to predict haplogroups in the 1,500 to 2,500 years which increases accuracy by 10X (90 % accuracy to 99 % accuracy). But you will always have 10 to 15 % of haplogroups with convergence with other haplogroups. So you either have to require 111 markers or report 80 to 90 % accuracy.

My newest version of YSNP prediction uses two variables (signature match number and genetic distance from signature). You can also do your own YSNP prediction for your own haplogroup project via EXCEL macros. All it takes is an easy to use $30 Statistical tool - AcaStat (great for generating the constants in the model) but really lacks on the accuracy reporting (expected for a $30 purchase tool where most tools cost $100 per month for rental). I have even documented how to implement YSNP predictor tools (these can also be converted to easy to understand empirical models that can visually been seen):

http://www.rcasey.net/DNA/R_L21/Math_behind_R_L21_SNP_Predictor_20180202A.pdf

rms2
06-02-2018, 02:17 PM
Interesting! Using my Ancestry data with Morley's predictor only gets me as far as R1b-L11.

RobertCasey
06-02-2018, 02:27 PM
rms2 - Morley's tool is not really a YSNP predictor. It searches your list of YSNPs and determines your terminal YSNP from a list of verified YSNP mutations. YSNP prediction uses YSTR markers and predicts YSNP branches. NevGen is another YSNP predictor tool but does not uses signatures but matches my L21 predictor tool (showing that there is always more than one mathematical way to arrive with the same results). My tool is based on signatures which most genetic genealogists are aware of. It allows you to visually see the math behind prediction. It can also be ported to EXCEL spreadsheets as well.

ArmandoR1b
06-02-2018, 09:49 PM
Interesting! Using my Ancestry data with Morley's predictor only gets me as far as R1b-L11.

You probably had the Ancestry v1 test. It did not test position 15654428 which is L21 in HG19/GRCh37. Ancestry v2 did test position 15654428. It is considered a prediction tool because there are errors with the chip based tests at times and the tool it isn't always 100% accurate. It also doesn't point out when well known subclades aren't tested so in your case it didn't tell you that L21 wasn't tested.

For 23andme raw files you can see which SNPs were tested by using the tool at http://www.y-str.org/2014/04/23andme-to-ysnps.html Use the Save Displayed Table. I use a spreadsheet that I created that basically does the same as that tool but also provides the longhand name of each SNP.

vettor
06-02-2018, 10:34 PM
Try uploading your data on Wegene. It also predicts Y DNA and mt DNA.

some times wegene finds it sometimes not

had trouble with cousins mtdna which wegene gave as R0
I used https://www.mitomap.org/mitomaster/index_snvs.cgi
and james lick .............Lick gave T1a mtdna and mitomaster gave T1a1e ..........they seem equal in qulity for mtdna

................................
reran my cousins ancestry results through morley and received


(most likely) I1

I1-L759 (I1-CTS10140, I1-CTS10338)

not knowning about I1 ......does this fall under I1-P109 branch?

I was given this by a relative I-Y4734* unsure anymore

ArmandoR1b
06-02-2018, 10:43 PM
You can still find 50 % of L21 haplogroups via my very dated L21 SNP predictor tool:

http://www.rcasey.net/DNA/R-L21_SNP_Predictor_Intro.html

Also, Nevgen and my tool combined reliably predicts around 60 to 70 %:

http://www.nevgen.org/

So 60 to 70 % of the time, these gateway SNP packs are not necessary. If somebody wants to get my tool current, we could easily get prediction up to 80 % coverage of R-L21. My predictor tool stopped when Big Y was introduced, so there have been a truckload of new predictable YSNPs that could be added. If you have the time, I would teach you how to continually add branches and I would provide the web site to host it. I also now use two parameters to predict haplogroups in the 1,500 to 2,500 years which increases accuracy by 10X (90 % accuracy to 99 % accuracy). But you will always have 10 to 15 % of haplogroups with convergence with other haplogroups. So you either have to require 111 markers or report 80 to 90 % accuracy.

My newest version of YSNP prediction uses two variables (signature match number and genetic distance from signature). You can also do your own YSNP prediction for your own haplogroup project via EXCEL macros. All it takes is an easy to use $30 Statistical tool - AcaStat (great for generating the constants in the model) but really lacks on the accuracy reporting (expected for a $30 purchase tool where most tools cost $100 per month for rental). I have even documented how to implement YSNP predictor tools (these can also be converted to easy to understand empirical models that can visually been seen):

http://www.rcasey.net/DNA/R_L21/Math_behind_R_L21_SNP_Predictor_20180202A.pdf

I think that Nqp15hhu did not have an STR test.

PoxVoldius
06-06-2018, 06:06 AM
Try uploading your data on Wegene. It also predicts Y DNA and mt DNA.

Does Ancestry DNA (V2 chip) even test enough mtDNA SNPs to determine a haplogroup?

I tried uploading both my data & my dad's data to Wegene, and for both of us on the mtDNA it only said L1'2'3'4'5'6.

Nqp15hhu
06-06-2018, 09:41 AM
Got RM269 on this, but not sure whether to trust that as the ancestry composition is terrible! Also it says that my mothers haplogroup is K1a1b2a1, which is common in Eastern Europe, so maybe I really do have Eastern European ancestry!

Tz85
06-06-2018, 01:57 PM
Got RM269 on this, but not sure whether to trust that as the ancestry composition is terrible! Also it says that my mothers haplogroup is K1a1b2a1, which is common in Eastern Europe, so maybe I really do have Eastern European ancestry!

K1a1b2a is common in Western Europe

Clarke
06-06-2018, 10:03 PM
I started a thread and then had to travel for a family emergency, so I'm sorry for not responding and I appreciate the responses. I will try responding properly later, just arrived back home and straight into work.

rms2
06-08-2018, 11:18 AM
Morley's tool for squeezing y-dna SNP's out of autosomal testing company data was a big help to me yesterday. 37-marker test results came in that all of us in our little project had been waiting for, expecting the test subject to match us and then perhaps to go on to do the Big Y, hopefully coming up with an as-yet-undiscovered SNP that would characterize the branch he and a supposed very close match with a fairly recent common ancestor share.

Well, that's not what we got. Not only did this gentleman, who shall remain anonymous, not match any of us, but he wasn't even in our y-dna haplogroup or even in a European haplogroup. FTDNA predicted E-M2 for him, which is a subclade downstream of E-V38 (E1b1a), a predominantly Subsaharan African haplogroup.

Given the man's paper trail, I thought perhaps there had been a mix-up or other error at FTDNA's lab. I didn't think it was likely, but one never knows. When I recalled that the man had already tested with Ancestry, I thought that running his raw data through Morley's tool would be a great way to check what FTDNA was saying. If Morley's tool produced a substantially different result, I could then approach FTDNA well prepared to demand a retest. So, I got this gentleman to send me his Ancestry raw data. When I ran it through Morley's tool, however, it produced the same result that FTDNA got, even predicting a terminal SNP of U174.

Then came the awkward task of explaining to a middle aged white guy the fact that he belongs to a Subsaharan African y-dna haplogroup and answering his questions about how that might have come to pass.

Anyway, Morley's tool saved me and him a lot of time we might have otherwise wasted calling FTDNA and asking them to double check, test his second sample, etc.

Titane
08-26-2018, 01:16 AM
Is it just me? I have all of a sudden noticed that AncestryDNA is giving yDNA and mtDNA haplogroup reports. I found it by accident at the bottom of the page in the black under « Other ». Of course I only have mtDNA, but the results are the same as those I got for a lot more money at FT-DNA.
So I need a male to check whether they see this or not.

PoxVoldius
08-26-2018, 01:34 AM
Is it just me? I have all of a sudden noticed that AncestryDNA is giving yDNA and mtDNA haplogroup reports. I found it by accident at the bottom of the page in the black under « Other ». Of course I only have mtDNA, but the results are the same as those I got for a lot more money at FT-DNA.
So I need a male to check whether they see this or not.

I administer my dad's kit (v2 chip) on AncestryDNA. But I'm not seeing this "Other" section or any link to any haplogroup reports. Is this on the smartphone app or the regular website? And which page of the site? Can you provide a screenshot?

I wonder if this is yet another thing that they're beta-testing without proper communication.

Titane
08-26-2018, 01:39 AM
I administer my dad's kit (v2 chip) on AncestryDNA. But I'm not seeing this "Other" section or any link to any haplogroup reports. Is this on the smartphone app or the regular website? And which page of the site? Can you provide a screenshot?

I wonder if this is yet another thing that they're beta-testing without proper communication.
It is all in French, on Ancestry.ca...25513
But last line is obvious I think.

PoxVoldius
08-26-2018, 01:45 AM
It is all in French, on Ancestry.ca...25513
But last line is obvious I think.

My list of links in that section, for all three of the kits that I administer, ends with the Sorenson (SMGF) link, and no links for haplogroup reports anywhere on the list. Same if I go in through Ancestry.ca instead of Ancestry.com; same if I switch it to the French language version of Ancestry.ca.

Must be yet another beta test that I'm not part of.

Titane
08-26-2018, 01:49 AM
My list of links in that section, for all three of the kits that I administer, ends with the Sorenson (SMGF) link, and no links for haplogroup reports anywhere on the list. Same if I go in through Ancestry.ca instead of Ancestry.com; same if I switch it to the French language version of Ancestry.ca.

Must be yet another beta test that I'm not part of.
Ok thanks...

Nqp15hhu
08-26-2018, 09:47 AM
I don’t have it! :(