PDA

View Full Version : New Y-DNA Nomenclature on 23andMe



Grossvater
11-16-2016, 04:05 AM
I got to looking at my "new 23andMe experience" results and realized they have abandoned the outdated Y-DNA nomenclature with something else. How up-to-date is the new nomenclature?

They have assigned me the new designation of R-PF6570. Does all this mean R1b-U152 or is there some new revelation in this designation?

Táltos
11-16-2016, 04:55 AM
I got to looking at my "new 23andMe experience" results and realized they have abandoned the outdated Y-DNA nomenclature with something else. How up-to-date is the new nomenclature?

They have assigned me the new designation of R-PF6570. Does all this mean R1b-U152 or is there some new revelation in this designation?

I believe it just means R1b-U152. My husband's changed to R-M529 aka R-L21.

wombatofthenorth
11-16-2016, 06:04 AM
Yeah, I think they just put this up today.
The new forums are broken today though.

sgdavies@hotmail.com
11-16-2016, 12:35 PM
Mine remain the same - in Germany.:\

Amerijoe
11-16-2016, 01:24 PM
R1a1a now R-M417.

sgdavies@hotmail.com
11-16-2016, 03:44 PM
R1a1a now R-M417.

They forgot to update mine, ahhh, I will wait also for confirmation on Haplogroups from my "Living DNA" results.

A Norfolk L-M20
11-16-2016, 05:19 PM
They haven't updated us because we are non-US customers, still on the "Old Experience". No sign of us ever receiving an "upgrade", although it was "soon coming" over a year ago.

ArmandoR1b
11-16-2016, 09:10 PM
It's frustrating that 23andme chose to use the least common synonym for SNPs such as PF6570 for U152, M529 for L21, and M412 for L51. They took forever to transition to shorthand SNP names but when they did they chose the ones people are less familiar with. It makes no sense.

Huitzilopochtli
11-16-2016, 09:54 PM
I have a (23andme 4th cousin) relative who was assigned I-P109, a fairly deep subclade of I1, while I'm still just I1-M253. Did anyone else get an increase in resolution?

ArmandoR1b
11-16-2016, 10:32 PM
I have a (23andme 4th cousin) relative who was assigned I-P109, a fairly deep subclade of I1, while I'm still just I1-M253. Did anyone else get an increase in resolution?

There shouldn't be an increase of resolution for anyone unless they were given the wrong haplogroup in the beginning. Since your cousin is now labeled as P109 then he should have been I1d1 before the change. You should have been just I1 before the change.

Huitzilopochtli
11-16-2016, 10:40 PM
There shouldn't be an increase of resolution for anyone unless they were given the wrong haplogroup in the beginning. Since your cousin is now labeled as P109 then he should have been I1d1 before the change. You should have been just I1 before the change.

Looking around more just now, I found another guy with I-L22. I had looked around before and I was unable to find anyone with a clade downstream of I1.

ArmandoR1b
11-16-2016, 10:48 PM
Looking around more just now, I found another guy with I-L22. I had looked around before and I was unable to find anyone with a clade downstream of I1.

Anyone that is now I-L22 was given I1d prior to the change. Everyone else that you are looking at that does not have a clade downstream of I1 is negative for the SNPs tested by 23andme that are downstream from M253. Those SNPs can be see at http://isogg.org/tree/2009/ISOGG_HapgrpI09.html

Leroy Jenkins
11-17-2016, 12:57 AM
My results are now less informative. Before the update, I was R-L2*. Now I'm R-PF6570 just like any other R-U152.

ArmandoR1b
11-17-2016, 01:25 PM
My results are now less informative. Before the update, I was R-L2*. Now I'm R-PF6570 just like any other R-U152.

That's because L2 had too high of a no-call rate. Even if you didn't have a no-call it seems they didn't want to cause confusion when two people have the same paternal line but a different haplogroup. I was R1b1b2a1a which is L52 but I had a no-call for that SNP so now I am M412 which is synonymous with L51 and was R1b1b2a1a. V2 had P312 but they took that out with V3. I don't know if that was a no-call issue with P312 or not but if there wasn't I wish they had still used it. I would at least show to be positive for that one.

jpb
11-17-2016, 11:29 PM
Went from C3e to C-M217.

leonardo
11-18-2016, 02:49 AM
Mine moved from R1a to R-M512. The site gives an age of 27,000 years. That can't be right.

ArmandoR1b
11-18-2016, 06:12 PM
Mine moved from R1a to R-M512.
That is funny. M512 rs17222146 position 16315153 isn't even tested by 23andme - https://you.23andme.com/tools/data/?query=rs17222146
Apart from that M512 is R1a1a and was even back in 2010.

R1a was defined by rs17307677 (L63) at 23andme - https://www.23andme.com/you/labs/haplogroup_tree_mut_mapper/results/?lineage_type=paternal&haplogroup=R1a

ISOGG 2016 has M511 and M513 as phylogenetic equivalents for R1a. http://isogg.org/tree/ISOGG_HapgrpR.html

M511 rs17307677 was tested by 23andme - https://you.23andme.com/tools/data/?query=rs17307677 but is not in the mutation mapper.


The site gives an age of 27,000 years. That can't be right.

They probably gave you the age of R1a according to the Poznik et al. (http://www.nature.com/ng/journal/v48/n6/full/ng.3559.html) calculation.

leonardo
11-18-2016, 07:51 PM
That is funny. M512 rs17222146 position 16315153 isn't even tested by 23andme - https://you.23andme.com/tools/data/?query=rs17222146
Apart from that M512 is R1a1a and was even back in 2010.

Wow. That is strange. So, it would not be much improvement for somebody. Fortunately, I have tested with the BigY.

Mike McG
11-19-2016, 02:45 PM
That's because L2 had too high of a no-call rate. Even if you didn't have a no-call it seems they didn't want to cause confusion when two people have the same paternal line but a different haplogroup. I was R1b1b2a1a which is L52 but I had a no-call for that SNP so now I am M412 which is synonymous with L51 and was R1b1b2a1a. V2 had P312 but they took that out with V3. I don't know if that was a no-call issue with P312 or not but if there wasn't I wish they had still used it. I would at least show to be positive for that one.

ArmandoR1b

I believe I was also R1b1b2a1a before the change and on the old site I was listed on the Mapper as:

Variant call anc der

rs13304168 (L52) T C T
rs9785659 (P311) G A G
rs9786076 (L11) C T C
rs9786283 (P310) C A C

I am now listed as P311, but a recent match, an Irish father (adopted) and son (2.47% & 2.09%, both 6 Seg.) who were also R1b1b2a1a, are now listed as M412. My father was Irish and I have tested DF27+ > Y5808+ and M756+. I thought they were probably also Y5808+ until they showed as M412+ rather than P311+. From your post it appears they could also possibly be P311+?

Mike

ArmandoR1b
11-19-2016, 07:19 PM
Wow. That is strange. So, it would not be much improvement for somebody. Fortunately, I have tested with the BigY.

That's good that you got a BigY test. I wish more people did SNP packs, SNP panels, BigY or FGC.

ArmandoR1b
11-19-2016, 07:29 PM
ArmandoR1b

I believe I was also R1b1b2a1a before the change and on the old site I was listed on the Mapper as:

Variant call anc der

rs13304168 (L52) T C T
rs9785659 (P311) G A G
rs9786076 (L11) C T C
rs9786283 (P310) C A C

I am now listed as P311, but a recent match, an Irish father (adopted) and son (2.47% & 2.09%, both 6 Seg.) who were also R1b1b2a1a, are now listed as M412. My father was Irish and I have tested DF27+ > Y5808+ and M756+. I thought they were probably also Y5808+ until they showed as M412+ rather than P311+. From your post it appears they could also possibly be P311+?

Mike

I am surprised that you were genotyped for P311, L11, and P310. Which version were you tested on? Yes, your recent matches are most likely also P311+ and DF27+ since they are negative for U106, U152, and L21 and DF27 is the most common of the remaining possible major subclades below M412.

Mike McG
11-20-2016, 05:32 PM
I am surprised that you were genotyped for P311, L11, and P310. Which version were you tested on? Yes, your recent matches are most likely also P311+ and DF27+ since they are negative for U106, U152, and L21 and DF27 is the most common of the remaining possible major subclades below M412.

ArmandoR1b

I ordered it to check my FTDNA results in Aug 2013. I believe it is version V3

Mike

ArmandoR1b
11-20-2016, 05:47 PM
ArmandoR1b

I ordered it to check my FTDNA results in Aug 2013. I believe it is version V3

Mike

Yes, that would be v3 according to the following 23andme page on chip versions (https://customercare.23andme.com/hc/en-us/articles/218392668-Understanding-the-different-versions-of-the-23andMe-genotyping-chip).

Luis
11-23-2016, 03:49 AM
The only thing they changed was the nomenclature, but they did not correct the wrong haplogroup assignments for people on V3. I tested with FTDNA and I'm R-P312 and so is my nephew, who took the 23andMe test when it was on V2. As a result, we naturally share the same haplogroup. However, because my brother was tested on V3, he was assigned a different haplogroup. Now he is said to be R-P311, which is wrong. We are both R-P312 and so is my nephew. When will they fix that?

ArmandoR1b
11-23-2016, 06:35 AM
The only thing they changed was the nomenclature, but they did not correct the wrong haplogroup assignments for people on V3. I tested with FTDNA and I'm R-P312 and so is my nephew, who took the 23andMe test when it was on V2. As a result, we naturally share the same haplogroup. However, because my brother was tested on V3, he was assigned a different haplogroup. Now he is said to be R-P311, which is wrong. We are both R-P312 and so is my nephew. When will they fix that?Probably never. You are most likely DF27 anyway so why care about P312 or P311? I would think that if you cared enough about your subclade that you would test at Yseq, FTDNA, or FGC and not care so much about 23andme which doesn't really care about our Y-DNA. 23andme mostly wants to make money off our DNA.

deadly77
11-26-2016, 09:39 PM
Looking around more just now, I found another guy with I-L22. I had looked around before and I was unable to find anyone with a clade downstream of I1.

I put my raw Y data from 23andme in the y-str.org analysis - the SNPs that 23andme tests for that are downstream of I-M253 are M227, M72, L22/S142, P109. I was negative for all of these so stayed at I1/I-M253.

ArmandoR1b
11-27-2016, 02:31 AM
Anyone that is now I-L22 was given I1d prior to the change. Everyone else that you are looking at that does not have a clade downstream of I1 is negative for the SNPs tested by 23andme that are downstream from M253. Those SNPs can be see at http://isogg.org/tree/2009/ISOGG_HapgrpI09.html


I put my raw Y data from 23andme in the y-str.org analysis - the SNPs that 23andme tests for that are downstream of I-M253 are M227, M72, L22/S142, P109. I was negative for all of these so stayed at I1/I-M253.

Correct, and those SNPs are found in the link that I had provided in a previous post that - http://isogg.org/tree/2009/ISOGG_HapgrpI09.html

deadly77
11-27-2016, 03:33 PM
Correct, and those SNPs are found in the link that I had provided in a previous post that - http://isogg.org/tree/2009/ISOGG_HapgrpI09.html

Also correct. My apologies - I hadn't looked at the link.

ddugas
12-16-2016, 09:25 PM
HI All:

Working on an adoption issue. We had an adoptee contact us. He is listed as R-P311 on 23andme. Based on everything he has told me I suspect he should match someone with a specific surname who is in 23andme (and related to him) but listed as R-M412. Based on everything I can read, P311 is a branch of M412 and there is a good chance these 2 are actually on the same twig but the different assignment is due to versions or no-calls?

Can anyone shed light? I don't know which versions each are using but the supposition I am making is that adoptee R-P311 is the son of a specific surname in the 23andme dataset which is listed as R-M412. Possible or have I missed something?

Thanks in advance,

ArmandoR1b
12-17-2016, 10:40 PM
HI All:

Working on an adoption issue. We had an adoptee contact us. He is listed as R-P311 on 23andme. Based on everything he has told me I suspect he should match someone with a specific surname who is in 23andme (and related to him) but listed as R-M412. Based on everything I can read, P311 is a branch of M412 and there is a good chance these 2 are actually on the same twig but the different assignment is due to versions or no-calls?

Can anyone shed light? I don't know which versions each are using but the supposition I am making is that adoptee R-P311 is the son of a specific surname in the 23andme dataset which is listed as R-M412. Possible or have I missed something?

Thanks in advance,

People that are positive for P311 and tested with v2 will have that as a terminal SNP and people that are positive for P311 but tested on v4 will have M412 as the terminal SNP. The only way to know for sure that they are from the same direct paternal line is for both of them to get at least a Y-DNA67 at FTDNA.

Ric
08-23-2018, 05:13 PM
It's frustrating that 23andme chose to use the least common synonym for SNPs such as PF6570 for U152, M529 for L21, and M412 for L51. They took forever to transition to shorthand SNP names but when they did they chose the ones people are less familiar with. It makes no sense.

It's more than unfortunate indeed, when you have 1000+ matches, you need filters to found interesting leads and since there is still no hierarchical organization, we need to know every single SNP tested as a key word for searching.
I wish there would be thread dedicated to those 23&me SNPs potentially indicative of df27.
In my experience, there are at least those :

L51 (many df27 individuals, including myself)
P312 (yes, one of my match is still P312 at 23&me while he's likely df27)
P311
Z209 (i got 3 matches)
CTS4065

what else ?

in the french forum, i mentioned the associated names

ArmandoR1b
08-27-2018, 11:58 AM
It's more than unfortunate indeed, when you have 1000+ matches, you need filters to found interesting leads and since there is still no hierarchical organization, we need to know every single SNP tested as a key word for searching.
I wish there would be thread dedicated to those 23&me SNPs potentially indicative of df27.
In my experience, there are at least those :

L51 (many df27 individuals, including myself)
P312 (yes, one of my match is still P312 at 23&me while he's likely df27)
P311
Z209 (i got 3 matches)
CTS4065

what else ?

in the french forum, i mentioned the associated names

It's important to keep in mind that different chips tested different SNPs and the update from several weeks? ago didn't affect customers prior to v5. Here are the dates of the release of the different chips and a little bit of information of what is and what isn't tested by the chips.

v1 Nov 2007 (P312 was not in ISOGG in 2007 so it wasn't included)
v2 Sept 2008 (tested for P312)
v3 Nov 2010 (Did not test for P312. Did test for P311. DF27 customers show as P311)
v4 Nov 2013 (Did not test for P312 or P311)
v5 Aug 2017 (Does not test for P312. Does test for P311 and several subclades of DF27)

Most people that show L51 as a terminal SNP at 23andme were tested by the v4 chip and since it does not test for P311 or P312 it is why they don't show positive for P311 or a downstream SNP.

Only the v2 chip tests for P312. Your match that is P312 was tested by the v2 chip.

The v5 chips tests for the following SNPs that are downstream from DF27. If there any missing, which would be extremely few, it is because they were not in the 2017 ISOGG tree or 23andme provides an untested phylogenetic equivalent to a tested SNP such as Z209 and Z220, respectively.

Z272
Z220/S356
Z295/S1217
Z278/S181
Z214/S348
CTS4065/S1221/Z2355
FGC15710/Y8717
DF17/S455
Z198/S228
S68/L165
CTS4188
L617
L881
A432
A431
Z225/F1343/S225

Anyone that gets Z209 has tested positive for the phylogenetic equivalent of Z220/S356 and given the terminal Z209 even though they weren't tested for Z209. This can be proven if any of your v5 matches that are Z209 are willing to go to the Raw Data Explorer page of rs750915391, which is the ID for Z209, at https://you.23andme.com/tools/data/?query=rs750915391 and report the result. It will show "not genotyped" The page for Z220 is at https://you.23andme.com/tools/data/?query=rs538725564 and they will have A under Your Genotype.

CTS4065 is at https://you.23andme.com/tools/data/?query=rs199595814

Another important point about the v5 chip is that it tests for D19, L238, and D99, so anyone that was tested by the v5 chip and terminal for P311 has a very high chance of being positive for DF27 if they are wiling to test for DF27 with another company.

By the way, I have way more R-L51 matches than any of the others. The v4 chip seems to have had many more customers than any of them so far. The v4 chip was used for almost 4 years and the v5 chip hasn't even been out for a year yet. Hopefully a v6 chip comes out soon with P312 and some more DF27 SNPs such as Z195 or Z196 and some more ZZ12 SNPs.

Ric
09-03-2018, 05:03 PM
It's important to keep in mind that different chips tested different SNPs and the update from several weeks? ago didn't affect customers prior to v5. Here are the dates of the release of the different chips and a little bit of information of what is and what isn't tested by the chips.

v1 Nov 2007 (P312 was not in ISOGG in 2007 so it wasn't included)
v2 Sept 2008 (tested for P312)
v3 Nov 2010 (Did not test for P312. Did test for P311. DF27 customers show as P311)
v4 Nov 2013 (Did not test for P312 or P311)
v5 Aug 2017 (Does not test for P312. Does test for P311 and several subclades of DF27)

Most people that show L51 as a terminal SNP at 23andme were tested by the v4 chip and since it does not test for P311 or P312 it is why they don't show positive for P311 or a downstream SNP.

Only the v2 chip tests for P312. Your match that is P312 was tested by the v2 chip.

The v5 chips tests for the following SNPs that are downstream from DF27. If there any missing, which would be extremely few, it is because they were not in the 2017 ISOGG tree or 23andme provides an untested phylogenetic equivalent to a tested SNP such as Z209 and Z220, respectively.

Z272
Z220/S356
Z295/S1217
Z278/S181
Z214/S348
CTS4065/S1221/Z2355
FGC15710/Y8717
DF17/S455
Z198/S228
S68/L165
CTS4188
L617
L881
A432
A431
Z225/F1343/S225

Anyone that gets Z209 has tested positive for the phylogenetic equivalent of Z220/S356 and given the terminal Z209 even though they weren't tested for Z209. This can be proven if any of your v5 matches that are Z209 are willing to go to the Raw Data Explorer page of rs750915391, which is the ID for Z209, at https://you.23andme.com/tools/data/?query=rs750915391 and report the result. It will show "not genotyped" The page for Z220 is at https://you.23andme.com/tools/data/?query=rs538725564 and they will have A under Your Genotype.

CTS4065 is at https://you.23andme.com/tools/data/?query=rs199595814

Another important point about the v5 chip is that it tests for D19, L238, and D99, so anyone that was tested by the v5 chip and terminal for P311 has a very high chance of being positive for DF27 if they are wiling to test for DF27 with another company.

By the way, I have way more R-L51 matches than any of the others. The v4 chip seems to have had many more customers than any of them so far. The v4 chip was used for almost 4 years and the v5 chip hasn't even been out for a year yet. Hopefully a v6 chip comes out soon with P312 and some more DF27 SNPs such as Z195 or Z196 and some more ZZ12 SNPs.

To be sure I understand, if you are labeled Z209 with the V5 chip, does it mean you are negative for CTS4065 (and downstearm) ?

ArmandoR1b
09-03-2018, 06:47 PM
To be sure I understand, if you are labeled Z209 with the V5 chip, does it mean you are negative for CTS4065 (and downstearm) ?
That is correct. Also, if a person is labeled as Z209 it is apparent that they were tested by the v5 chip.

M167/SRY2627 and M153 are the only SNPs downstream from DF27 that were tested by previous chips. M167/SRY2627 should have been in the list above for v5 (It was an error due to an attempt to remove SNPs with duplicate rsids). M153 is not tested by v5.

Ric
09-03-2018, 07:31 PM
Also, if a person is labeled as v5 it is apparent that they were tested by the v5 chip.that info does not appear at 23&me as far as I can see. Perhaps in the 'scientific' sub-section ?
Only at gedmatch people may be labeled 'V5'.

This is a bit disappointing since as you mentioned above, most 23&me customers have tested on V4, perhaps as much as 2 millions people, and they are not going to re-test just to refine their haplogroup. Now this company is facing a lot of competition and the V5 customer base may never reach the level of the V4.

ArmandoR1b
09-03-2018, 08:36 PM
that info does not appear at 23&me as far as I can see. Perhaps in the 'scientific' sub-section ?
Only at gedmatch people may be labeled 'V5'.

This is a bit disappointing since as you mentioned above, most 23&me customers have tested on V4, perhaps as much as 2 millions people, and they are not going to re-test just to refine their haplogroup. Now this company is facing a lot of competition and the V5 customer base may never reach the level of the V4.

I corrected that to what I meant to write. It should have been "Also, if a person is labeled as Z209 it is apparent that they were tested by the v5 chip." Sorry about that error.

The chip version can be checked at https://you.23andme.com/user/ next to Genotyping Chip Version.

It will be a long time until the v5 customer count reaches the same number as the v4 customer count but I think that it will happen. 23andme is still a popular product. I wish AncestryDNA and myHeritage weren't so popular since their ethnicity calculators aren't as good as the one by 23andme and they don't test as many haplogroups as 23andme.

geebee
09-04-2018, 12:44 AM
I corrected that to what I meant to write. It should have been "Also, if a person is labeled as Z209 it is apparent that they were tested by the v5 chip." Sorry about that error.

The chip version can be checked at https://you.23andme.com/user/ next to Genotyping Chip Version.

It will be a long time until the v5 customer count reaches the same number as the v4 customer count but I think that it will happen. 23andme is still a popular product. I wish AncestryDNA and myHeritage weren't so popular since their ethnicity calculators aren't as good as the one by 23andme and they don't test as many haplogroups as 23andme.

An ethnicity calculator is just one reason to buy a DNA test. Some of us are interested in building and proving our family trees. For that purpose, 23andMe v5 is actually a pretty lousy test. The reason is, very poor compatibility with other tests -- including, in fact, 23andMe's own v2, v3, and v4.

Ric
09-04-2018, 12:58 AM
The reason is, very poor compatibility with other tests -- including, in fact, 23andMe's own v2, v3, and v4.can you elaborate on that ?

ArmandoR1b
09-04-2018, 12:48 PM
An ethnicity calculator is just one reason to buy a DNA test. Some of us are interested in building and proving our family trees. For that purpose, 23andMe v5 is actually a pretty lousy test. The reason is, very poor compatibility with other tests -- including, in fact, 23andMe's own v2, v3, and v4.

The DNA test at AncestryDNA is not needed to build a family tree of known relatives at Ancestry.com. In the case of people wanting to prove relationships through DNA testing then yes the v5 test has a problem since the raw data can't be used with regular Gedmatch. However, in my experience, and what I have read on various forums, is that a majority of people are not very interested in proving relationships. They are more interested in finding or proving their ethnicity. It is that majority of people, that don't care so much about finding relatives that I wish would choose 23andme over AncestryDNA. Anyone that wants to find as many relatives as possible should be testing with all of the DNA companies and not just with one since there are people that have tested with 23andme that have never tested with AncestryDNA. Really I would be happy with 23andme being equally popular as AncestryDNA.

ArmandoR1b
09-04-2018, 12:57 PM
can you elaborate on that ?

There is not a lot of overlap between the positions tested by the v5 Illumina GSA chip and the positions tested by the previous 23andme chips or with the chips used by FTDNA, AncestryDNA, and myHeritage. It is also the reason that the v5 raw DNA can't be used with regular Gedmatch and it is why Gedmatch Genesis (https://genesis.gedmatch.com) was created.

23andme uses imputation to determine some of the calls of the non-overlapping positions. According to Illumina that imputation is highly accurate.

The LivingDNA chip uses a customized version of the Illumina GSA chip used by 23andme v5 so the raw data isn't compatible with regular Gedmatch either and Gedmatch Genesis has to be used.

geebee
09-04-2018, 02:12 PM
The DNA test at AncestryDNA is not needed to build a family tree of known relatives at Ancestry.com.

For some people, yes. If there are no discrepancies between what they have been told -- or even what has been placed in records -- then you can build a tree even without DNA testing. But, what people are told is not always true; and even records can be false. For example, a birth record generally contains whatever information was supplied by the mother -- and her alone.

In the case of my maternal grandfather, his birth record actually contains information supplied by himself, when he was 59 years old. Presumably, it's based on what his parents told him, and it's supported by affidavits from his two older brothers. I have reason to believe it's also false.

What sort of reason? DNA evidence. Either the person I knew as my grandfather was not my mother's biological father, or the couple who raised him were not his biological parents. I believe the latter is true, but whether the latter or the former is true, I have a pretty solid basis in DNA evidence for believing that I know the identity of my mother's paternal grandparents. The only question is, was the man who raised her their son?

I won't go into detail here about why I say this, because it's tangential to my point. Which, again, is just that what people think they know about their ancestry can be wrong, even if there is documentation to back it up. DNA evidence should be believed over documentation, which can be the product of people who either didn't actually know or had reason to lie.

ArmandoR1b
09-04-2018, 02:38 PM
For some people, yes. If there are no discrepancies between what they have been told -- or even what has been placed in records -- then you can build a tree even without DNA testing. But, what people are told is not always true; and even records can be false. For example, a birth record generally contains whatever information was supplied by the mother -- and her alone.

I knew what you were getting at and why I had mentioned that "In the case of people wanting to prove relationships through DNA testing then yes the v5 test has a problem since the raw data can't be used with regular Gedmatch" That should have been included in the quote. Most people aren't into DNA testing enough to go through all of that trouble just prove relationships. That isn't why most people get a DNA test at AncestryDNA. If everyone that gets an AncestryDNA test were to also get a 23andme test then it would help get more people interested in Y-DNA testing.

In my own case, I have proven relationships in my tree with DNA testing from FTDNA and 23andme. I haven't needed AncestryDNA testing to prove them.

geebee
09-04-2018, 06:07 PM
can you elaborate on that ?

ArmandoR1b has already elaborated on this, and for the most part I'd say the same thing. However, in response to his statement that


23andme uses imputation to determine some of the calls of the non-overlapping positions. According to Illumina that imputation is highly accurate

the v5 chip is in fact an Illumina product. (As are earlier versions, as well.) Not everyone agrees that the "imputation is highly accurate", but in any case it is important to know this is an example of a manufacturer essentially endorsing its own product.

geebee
09-04-2018, 06:47 PM
[QUOTE=ArmandoR1b;479209]I knew what you were getting at and why I had mentioned that "In the case of people wanting to prove relationships through DNA testing then yes the v5 test has a problem since the raw data can't be used with regular Gedmatch" That should have been included in the quote. Most people aren't into DNA testing enough to go through all of that trouble just prove relationships. That isn't why most people get a DNA test at AncestryDNA. If everyone that gets an AncestryDNA test were to also get a 23andme test then it would help get more people interested in Y-DNA testing.

But that's just it. I wasn't originally someone "wanting to prove relationships through DNA testing". That's why I didn't include that part of the quote -- because what I was testing for wasn't actually relevant to the result.

It was as a result of testing that I learned 1/4 of my entire tree was false. It wasn't that I tested in order to "prove" my tree. Fortunately, I did my first autosomal test at 23andMe (on v2, which I later upgraded to v3). It was sometime afterward that I encountered the first of two 2nd cousins (and one 2nd cousin once removed) who caused me to suspect that either (1) my maternal grandfather's parents weren't who they'd been claimed to be; or (3) my grandfather himself was not who we thought he was.

So, yes, a person not interested in "proving" his or her tree can go ahead and test on v5, but they should realize that it will be more difficult to use that test for comparison to relatives tested on any other chip. While they may not initially care about that, it will be equally true later if they decide they do care.

Ric
09-09-2018, 06:01 PM
The v5 chips tests for the following SNPs that are downstream from DF27. If there any missing, which would be extremely few, it is because they were not in the 2017 ISOGG tree or 23andme provides an untested phylogenetic equivalent to a tested SNP such as Z209 and Z220, respectively.

Z272
Z220/S356
Z295/S1217
Z278/S181
Z214/S348
CTS4065/S1221/Z2355
FGC15710/Y8717
DF17/S455
Z198/S228
S68/L165
CTS4188
L617
L881
A432
A431
Z225/F1343/S225

Anyone that gets Z209 has tested positive for the phylogenetic equivalent of Z220/S356 and given the terminal Z209 even though they weren't tested for Z209. This can be proven if any of your v5 matches that are Z209 are willing to go to the Raw Data Explorer page of rs750915391, which is the ID for Z209, at https://you.23andme.com/tools/data/?query=rs750915391 and report the result. It will show "not genotyped" The page for Z220 is at https://you.23andme.com/tools/data/?query=rs538725564 and they will have A under Your Genotype.

If I may come back on this issue, I understand what you said above, but I assume that the V5 chip doesn't always give 100% certainty at all time for all SNPs. A poor read would be the equivalent of a 'low confidence' call. And such 'low confidence' call for CTS4065 combined with other SNPs may induce the 23&me algorithms to label the person Z209 instead of CTS4065.
Is it possible and what would be the frequency of such poor read/low confidence ?

If as much as 30% CTS4065 individuals were read with 'low confidence' for this SNPs by the V5 chip, and labeled as z209 instead, then about one third of my three Z209 matches could potentially be CTS4065, which means one of them is actually a relatively close paternal relative.


Is there anybody here who is a known CTS4065+ at FTDNA, but nonetheless labeled Z209 at 23&me ?

hmm I guess there is not enough V5 people yet to answer that. But based on 1 sample, my neighbor BY12176 downstream of CTS4065, the success rate for reading this SNP is 100% since he is correctly labeled as such at 23&me. 23&me should propose some sort of an update, or a partial retest for the Y haplogroup, for the V4 people like me who are stuck at the ridiculously high L51 level.

ArmandoR1b
09-09-2018, 07:24 PM
If I may come back on this issue, I understand what you said above, but I assume that the V5 chip doesn't always give 100% certainty at all time for all SNPs. A poor read would be the equivalent of a 'low confidence' call. And such 'low confidence' call for CTS4065 combined with other SNPs may induce the 23&me algorithms to label the person Z209 instead of CTS4065.
Is it possible and what would be the frequency of such poor read/low confidence ?

The test is almost 100% reliable. I've never seen anyone with "not determined" (which is how no-calls are labeled nowadays in the Raw Data Explorer) for a Y-DNA SNP downstream from a major haplogroup. CTS4065 is rs199595814 defined by hg19 grch37 position 15372323 mutation of A to C. You can ask your Z209 matches to check their genotype for CTS4065 by going to https://you.23andme.com/tools/data/?query=rs199595814 and looking under Your Genotype. They will all have A meaning negative for CTS4065. Everyone tested by v4 or earlier will have "not genotyped" which means not tested.




If as much as 30% CTS4065 individuals were read with 'low confidence' for this SNPs by the V5 chip, and labeled as z209 instead, then about one third of my three Z209 matches could potentially be CTS4065, which means one of them is actually a relatively close paternal relative.
If there are any no-calls for CTS4065 it will be a very rare occurrence. Most likely less than 1% if there are.



Is there anybody here who is a known CTS4065+ at FTDNA, but nonetheless labeled Z209 at 23&me ?
I have no idea but I strongly doubt it.



hmm I guess there is not enough V5 people yet to answer that. But based on 1 sample, my neighbor BY12176 downstream of CTS4065, the success rate for reading this SNP is 100% since he is correctly labeled as such at 23&me. 23&me should propose some sort of an update, or a partial retest for the Y haplogroup, for the V4 people like me who are stuck at the ridiculously high L51 level.
Supposedly 23andme has talked about something like that but I have no idea how serious of a consideration it is. They almost definitely would charge for it since that would be an extra expense for them to spend all of the time and materials on new testing.

Ric
09-09-2018, 07:56 PM
Supposedly 23andme has talked about something like that but I have no idea how serious of a consideration it is. They almost definitely would charge for it since that would be an extra expense for them to spend all of the time and materials on new testing.

If it is a partial test 'a la carte' for the Y chromosome, the Mit, or something else, assuming 23&me has kept the DNA samples, paying a fee is understandable. But if they have to resend an entire kit again, it's not an upgrade and that would be less attractive in my opinion.

JonikW
09-15-2018, 11:10 PM
I've just found my late mum's first cousin on Gedmatch, who I didn't even know existed, but he hasn't tested Y DNA as far as I can work out from what he's said so far. My dream was always to find out my mum's Jones paternal haplogroup, which is this line. My question is, if, for example, her family turned out to be L21, as is likely for Wales, how far down the line would v5 get them?

ArmandoR1b
09-16-2018, 12:00 AM
I've just found my late mum's first cousin on Gedmatch, who I didn't even know existed, but he hasn't tested Y DNA as far as I can work out from what he's said so far. My dream was always to find out my mum's Jones paternal haplogroup, which is this line. My question is, if, for example, her family turned out to be L21, as is likely for Wales, how far down the line would v5 get them?

If he tested with AncestryDNA or myHeritage then he was tested for L21 and some subclades of L21. They are a lot fewer subclades of L21 than tested by 23andme v5 but at least L21 is tested. You can upload his raw data to https://ytree.morleydna.com/extractFromAutosomal to see which SNPs he shows positive for. Then if you want a more downstream subclade then get a Yseq or FTDNA BigY test for him.

Ric
10-12-2018, 12:28 PM
It's important to keep in mind that different chips tested different SNPs and the update from several weeks? ago didn't affect customers prior to v5. Here are the dates of the release of the different chips and a little bit of information of what is and what isn't tested by the chips.

v1 Nov 2007 (P312 was not in ISOGG in 2007 so it wasn't included)
v2 Sept 2008 (tested for P312)
v3 Nov 2010 (Did not test for P312. Did test for P311. DF27 customers show as P311)
v4 Nov 2013 (Did not test for P312 or P311)
v5 Aug 2017 (Does not test for P312. Does test for P311 and several subclades of DF27)

Most people that show L51 as a terminal SNP at 23andme were tested by the v4 chip and since it does not test for P311 or P312 it is why they don't show positive for P311 or a downstream SNP.

Only the v2 chip tests for P312. Your match that is P312 was tested by the v2 chip.

The v5 chips tests for the following SNPs that are downstream from DF27. If there any missing, which would be extremely few, it is because they were not in the 2017 ISOGG tree or 23andme provides an untested phylogenetic equivalent to a tested SNP such as Z209 and Z220, respectively.

Z272
Z220/S356
Z295/S1217
Z278/S181
Z214/S348
CTS4065/S1221/Z2355
FGC15710/Y8717
DF17/S455
Z198/S228
S68/L165
CTS4188
L617
L881
A432
A431
Z225/F1343/S225

Anyone that gets Z209 has tested positive for the phylogenetic equivalent of Z220/S356 and given the terminal Z209 even though they weren't tested for Z209. This can be proven if any of your v5 matches that are Z209 are willing to go to the Raw Data Explorer page of rs750915391, which is the ID for Z209, at https://you.23andme.com/tools/data/?query=rs750915391 and report the result. It will show "not genotyped" The page for Z220 is at https://you.23andme.com/tools/data/?query=rs538725564 and they will have A under Your Genotype.

CTS4065 is at https://you.23andme.com/tools/data/?query=rs199595814

Another important point about the v5 chip is that it tests for D19, L238, and D99, so anyone that was tested by the v5 chip and terminal for P311 has a very high chance of being positive for DF27 if they are wiling to test for DF27 with another company.

By the way, I have way more R-L51 matches than any of the others. The v4 chip seems to have had many more customers than any of them so far. The v4 chip was used for almost 4 years and the v5 chip hasn't even been out for a year yet. Hopefully a v6 chip comes out soon with P312 and some more DF27 SNPs such as Z195 or Z196 and some more ZZ12 SNPs.

Is this list of SNPs tested by 23&me V5 still up to date ?
Is there any hints that previous chip version (up to V4) can sometimes be relabeled to downstream SNPs, beside those indicated in the quote above ?

This thread may have the most practical benefits for customers that have both tested at FTDNA-SNPs and 23&me. For example, I just got a new Z-214 match at 23&me, therefore df27+. He is Anonymous, but our common matches suggest he is of Spanish ancestry. Unfortunately I still haven't got any CTS4065+ match. But it is good to see that some DF27 branches show up at at 23&me.

Also, one of my Z209 autosomal match at 23&me vanished recently. Or alternatively he switched to anonymous while his haplogroup was downgraded to P311 (possible ?). I don't know what's going on with 23&me. My number of matches oscillates around 1000, and most of the time I can't figure out who is coming and who's leaving my list.

ArmandoR1b
10-15-2018, 02:45 PM
Is this list of SNPs tested by 23&me V5 still up to date ?
Is there any hints that previous chip version (up to V4) can sometimes be relabeled to downstream SNPs, beside those indicated in the quote above ?

This thread may have the most practical benefits for customers that have both tested at FTDNA-SNPs and 23&me. For example, I just got a new Z-214 match at 23&me, therefore df27+. He is Anonymous, but our common matches suggest he is of Spanish ancestry. Unfortunately I still haven't got any CTS4065+ match. But it is good to see that some DF27 branches show up at at 23&me.

Also, one of my Z209 autosomal match at 23&me vanished recently. Or alternatively he switched to anonymous while his haplogroup was downgraded to P311 (possible ?). I don't know what's going on with 23&me. My number of matches oscillates around 1000, and most of the time I can't figure out who is coming and who's leaving my list.

I have found out that Z272 and Z295 are always negative even for people that are positive for downstream SNPs so they are not reliable SNPs.

As far as my matches the following DF27 SNPs tested by the v5 chip are showing up -

R-Z209 (not tested. It is actually Z220 that is tested)
R-Z278
R-Z214
R-CTS4065
R-FGC15710
R-M167
R-CTS4188
R-Z225


SRY2627/M167 is also tested by the v5 chip. I mentioned that one in post #34

Ric
11-03-2018, 02:45 PM
I have found out that Z272 and Z295 are always negative even for people that are positive for downstream SNPs so they are not reliable SNPs.

As far as my matches the following DF27 SNPs tested by the v5 chip are showing up -

R-Z209 (not tested. It is actually Z220 that is tested)
R-Z278
R-Z214
R-CTS4065
R-FGC15710
R-M167
R-CTS4188
R-Z225


SRY2627/M167 is also tested by the v5 chip. I mentioned that one in post #34

Hi Armando

Are you sure CTS4065 is 100% reliable at 23&me ?
I just got a new DNA-relative V5 who has my name anglicized and the guy is indicated as Z209. Which makes me think that we are closely related in the paternal line, enough to be autosomal match anyways, and that I should therefore be labeled Z209 as well if I was in V5. It is of course impossible to say for sure what would be my haplogroup if I was V5. It should be CTS4065 with V5 according to what you said above, but....this guy makes me doubt.
As I mentioned above, my closest neighbor in FTDNA haplotree is labeled CTS4065 at 23&me, but we diverged more than 30SNPs ago (total) and I wonder if these not-shared variants could cause me to appear negative for CTS4065 at 23&me.

ArmandoR1b
11-03-2018, 06:19 PM
Hi Armando

Are you sure CTS4065 is 100% reliable at 23&me ?
I just got a new DNA-relative V5 who has my name anglicized and the guy is indicated as Z209. Which makes me think that we are closely related in the paternal line, enough to be autosomal match anyways, and that I should therefore be labeled Z209 as well if I was in V5. It is of course impossible to say for sure what would be my haplogroup if I was V5. It should be CTS4065 with V5 according to what you said above, but....this guy makes me doubt.
As I mentioned above, my closest neighbor in FTDNA haplotree is labeled CTS4065 at 23&me, but we diverged more than 30SNPs ago (total) and I wonder if these not-shared variants could cause me to appear negative for CTS4065 at 23&me.

There are too many people with matches that are positive for CTS4065 for it not to be a highly reliable. Your autosomal match most likely belongs to a different subclade of Z209 than people that are positive for CTS4065. It's expected that people from the same gene pool to have autosomal matches with haplogroups of similar older branches but different subclades. That's how population genetics works. There is always a chance of a one-off for an SNP through the 23andme test but other than a situation like that you would show to be positive for CTS4065 at 23andme through the v5 test.

ADW_1981
11-03-2018, 08:40 PM
Recently, a descendant of my maternal great grandfather appeared in my matches at 23andMe. He is labeled as P311 on whatever this latest chip is. My original results were on v2 and labeled me as P312 (even though im below Z209), and my v3 results labeled me as P311. What is the consensus on the latest chip? Do they yield for P312, DF27, L21 and the like? Or do all groups of P312 get labeled as P311?

Thanks in advance.

ArmandoR1b
11-04-2018, 01:05 PM
Recently, a descendant of my maternal great grandfather appeared in my matches at 23andMe. He is labeled as P311 on whatever this latest chip is. My original results were on v2 and labeled me as P312 (even though im below Z209), and my v3 results labeled me as P311. What is the consensus on the latest chip? Do they yield for P312, DF27, L21 and the like? Or do all groups of P312 get labeled as P311?

Thanks in advance.

The 23andme v5 chip does not test for P312, DF27, Z195 or Z196.

It does test for U106, L21, U152, DF19, DF99, and L238 plus all of the the DF27 SNPs I've previous listed:

R-Z209 (not tested. It is actually Z220 that is tested)
R-Z278
R-Z214
R-CTS4065
R-FGC15710
R-M167
R-CTS4188
R-Z225
SRY2627/M167

artemv
11-04-2018, 09:31 PM
I can see you all discuss the R1b downstream SNP's covered by 23andMe tests.
Maybe somebody knows a little bit about downstream SNP's of other Y-happlogroups?

I know that v4 23andMe tests L161. In "paternal happlogroup" they only show I2-M423, but it is possible to check if you are L161+ or L161-. This gives more data for everyone tested by 23andMe as M423, because branches of M423 other than L161 and L621 are very rare.
Does anyone know anything about other downstream SNP's of I2? I tried to search, but did not find anything...

ArmandoR1b
11-04-2018, 09:53 PM
I can see you all discuss the R1b downstream SNP's covered by 23andMe tests.
Maybe somebody knows a little bit about downstream SNP's of other Y-happlogroups?

I know that v4 23andMe tests L161. In "paternal happlogroup" they only show I2-M423, but it is possible to check if you are L161+ or L161-. This gives more data for everyone tested by 23andMe as M423, because branches of M423 other than L161 and L621 are very rare.
Does anyone know anything about other downstream SNP's of I2? I tried to search, but did not find anything...

See https://tinyurl.com/y8baooht for some of the haplogroups that 23andme reports. Some of those are from older versions though. 23andme sometimes uses a different phylogenetic equivalent than the one that was actually tested.

See https://tinyurl.com/y94td7q3 for a table with the 2017 longhand names of the rsids that are in the raw data. These are the ones that are actually tested.

artemv
11-05-2018, 11:55 PM
See https://tinyurl.com/y8baooht for some of the haplogroups that 23andme reports. Some of those are from older versions though. 23andme sometimes uses a different phylogenetic equivalent than the one that was actually tested.

See https://tinyurl.com/y94td7q3 for a table with the 2017 longhand names of the rsids that are in the raw data. These are the ones that are actually tested.

Thank you.
I was tested on v4, so these tables do not really help in my case - I have checked that I do not have any positions ineresting for me.

Ric
11-06-2018, 12:30 AM
Thank you.
I was tested on v4, so these tables do not really help in my case - I have checked that I do not have any positions ineresting for me.

Obviously 23&me is not gonna retest, so if you are V4 like me or older, we are going to keep our old haplogroup. I recall that 23&me sent an email to all its customers 2 or 3 years ago, because they pass 'a million customers' or so. So V1+V2 + V3 + V4 are probably 80% of 23&me database. As I said that email was 2 years ago, so how much is it now, 2 millions ? Unfortunatly I think 23&me' user base is not going to grow as fast as it did during these early versions, because there are lots of competing companies now. Many french people don't go on 23&me anymore as they used to.

Perhaps a re-run of the algorithms for the Y haplogroup could be used to find SNPs that were tested by the older chip version (V4 and before) perhaps some of these SNPs 'correlate' with the SNPs tested by the new V5 version. For example as Armando mentioned, Z209 is not tested in V5, but Z220 is, and V5 algorithms call Z209 when Z220 is positive, so perhaps something similar could be done for V4 and older... so that people with a very shallow haplogroup like mine, or yours I guess, could enjoy some improvement. That would be a 'software' re-run of course, no way they could re-test the DNA samples of 2 millions customers on their new chips, assuming they kept the DNA. But even a software re-run for 2 millions customers, ouch, that can be free, but I'd pay a fee for it.

ArmandoR1b
11-06-2018, 02:28 PM
Perhaps a re-run of the algorithms for the Y haplogroup could be used to find SNPs that were tested by the older chip version (V4 and before) perhaps some of these SNPs 'correlate' with the SNPs tested by the new V5 version. For example as Armando mentioned, Z209 is not tested in V5, but Z220 is, and V5 algorithms call Z209 when Z220 is positive, so perhaps something similar could be done for V4 and older... so that people with a very shallow haplogroup like mine, or yours I guess, could enjoy some improvement.

There aren't any Y-chromosome positions tested by the v4 chip that would provide a different haplogroup name than what is already given to v4 customers. For instance, in 2017, 23andme provided R-U106 v4 customers with the paternal haplogroup of R-M405 but now it is R-U106. That's the end of the road for v4 testers since if they had tested positive for something downstream they would have been given the pertinent haplogroup such as R-M467 for R-U198.

In the v5 test there are less than 25 phylogenetic equivalent SNPs used as paternal haplogroups that are not tested. Z209 is one of those uncommon cases. Most are tested and the name change was done between synonymous SNPs.

Ric
11-06-2018, 03:16 PM
That's the end of the road for v4 testers since if they had tested positive for something downstream they would have been given the pertinent haplogroup such as R-M467 for R-U198.

You are crushing my hopes. I have so many 'L51' matches, some of them could be very close to my haplogroup.

There is something else strange, I noticed that 23&me now describes L51 as a 'rare group' and it indicates R-L51 is 1/7400 customer base !
How can that be ? L51 is a major haplogroup among western European men, who constituted the initial 23&me customer base. So unless western men have become a minority among all 23&me customers, this is misleading.

ArmandoR1b
11-06-2018, 03:39 PM
Thank you.
I was tested on v4, so these tables do not really help in my case - I have checked that I do not have any positions ineresting for me.

I wasn't sure if you wanted v5 data or not. Some people are interested in what is tested in v5 that isn't tested in v4. If you are still interested in the I2 SNPs tested by the v4 that are in the 2018 ISOGG SNP Index you can check the table at - https://tinyurl.com/ydywdl92

ArmandoR1b
11-06-2018, 03:50 PM
You are crushing my hopes. I have so many 'L51' matches, some of them could be very close to my haplogroup.

There is something else strange, I noticed that 23&me now describes L51 as a 'rare group' and it indicates R-L51 is 1/7400 customer base !
How can that be ? L51 is a major haplogroup among western European men, who constituted the initial 23&me customer base. So unless western men have become a minority among all 23&me customers, this is misleading.

I agree that it is misleading. They should provide more information on how they are getting those stats and if they aren't using stats of just v4 customers for our stats then they need inform customers that they weren't tested for many other SNPs that v5 customers were tested for.

ADW_1981
11-26-2018, 03:05 AM
I have contacted a descendant of my maternal gg-father (or his lineage) and found it to be R-P311 on the v5 chip. Are there any more informative SNPs that can be yielded from the raw data without ordering through another service other than 23andMe? (through a 3rd party tool)

ArmandoR1b
11-26-2018, 05:53 AM
I have contacted a descendant of my maternal gg-father (or his lineage) and found it to be R-P311 on the v5 chip. Are there any more informative SNPs that can be yielded from the raw data without ordering through another service other than 23andMe? (through a 3rd party tool)

23andme v5 reports the most downstream positive SNP out of all of the SNPs tested. The last update they did fixed the problem of some positive SNPs not being used for the most downstream SNP.

You can see a spreadsheet of all of the positions that are included in the report at https://tinyurl.com/y7fyu65r It has the 2017 longhand names sorted alphabetically so it makes it easier to figure out what is downstream if you download the spreadsheet and set the filter to R1b1a1a2a1a to only show P311 and downstream SNPs.

If you download the program at https://tinyurl.com/yakurs26 and run the raw data file through it you will see all of the downstream SNPs, that aren't false positives, have the negative allele. The program is an improved version of a tool created by Felix. You can save the displayed table as a CSV and open it with Excel or Google Sheets.

The best thing to do is to ask him to get the R1b-M343 Orientation Panel (https://www.yseq.net/product_info.php?cPath=27&products_id=16349) from Yseq in order to determine if he is positive for any of the SNPs included in that panel not tested by 23andme.