R* haplotype designation? (23andme)

[khanabadoshi]

I have tested a friend of mine and he was assigned an interesting y haplogroup designation -- R -- via 23andMe. His father is a Rajput and his mother, Kashmiri.




Naturally, I was curious. I took his 23andMe raw data and used, Felix Immanuel's 23andMe to Y-SNP Converter (2014) and input the data into Chris Morley's Y-SNP subclade predictor, optimized for Geno 2.0 data transferred to FTDNA, and also the Full Genomes demo data. Even though it isn't meant for 23andMe data, I did get results -- though based on the 2013 tree:




The following SNPs were NOT accounted for:





I was hoping to get some input and interpretation from you good folks.

A link to his y-SNPs, RSID, and position in .csv format.
A link to his y-SNPs in a text document.

(Both of these files were generated via the 23andMe to Y-SNP Converter.)

[Huntergatherer1066]

Those are certainly some interesting results; I would hope he considers further testing, preferably an NGS test to confirm and further test his Y-DNA. Verifying he is really P227- and P285- via Sanger sequencing might also be interesting; that would be a new initial fork in R. I am wary though with it being a chip test optimized especially for autosomal testing, not Y-DNA, which is why I think verifying through another testing means is important (note the most likely false negative in haplogroup F for CTS3868 for instance).

[khanabadoshi]

Those are certainly some interesting results; I would hope he considers further testing, preferably an NGS test to confirm and further test his Y-DNA. Verifying he is really P227- and P285- via Sanger sequencing might also be interesting; that would be a new initial fork in R. I am wary though with it being a chip test optimized especially for autosomal testing, not Y-DNA, which is why I think verifying through another testing means is important (note the most likely false negative in haplogroup F for CTS3868 for instance).

I think I will have to convince him to do so, his results require clarification. It's been suggested to me that most people who get an R* result are usually designated as R2 when further tested? (I might be mistaken in my recall).

Maybe some experts will glance over the SNPs in the attached links and find something meaningful. If someone requires the raw data to reach some conclusions, let me know!

[Megalophias]

Sweet.

An R*(xR1, R2) has been previously reported in a Kyrgyz man, but that didn't test any R SNPs other than M207.

[Huntergatherer1066]

I think I will have to convince him to do so, his results require clarification. It's been suggested to me that most people who get an R* result are usually designated as R2 when further tested? (I might be mistaken in my recall).

Maybe some experts will glance over the SNPs in the attached links and find something meaningful. If someone requires the raw data to reach some conclusions, let me know!

R2* seems to be a likely outcome if P227 and P285 are false negatives, which seems likely. He could test for an R2 SNP like M479 I guess. R2* would still be very interesting.

[Huntergatherer1066]

I looked at the raw data file and for P285 it says he is not derived and that his genotype is A, however P285 is a C>A mutation. I think there may be confusion there because there is another SNP at the same location PF6059 listed as a T to C mutation.

I looked at my paternal grandfather's 23andMe results (he is R1b-Z2534) and for P227 they also have him as G/ancestral so they must just not be scoring it properly. I couldn't find P285 but he was tested with the V3 chip which might not have it.

[DMXX]

As mentioned in private correspondence recently, much of the supposed cases of R*-M207 found through FTDNA samples from South-Central Asia have turned out R2* following the discovery of the M479 SNP by Myres et al. - Vineviz's Y-DNA R* FTDNA Project demonstrates this. As such, older studies with samples which appear to be R*-M207 (such as Sharma et al.) cannot reliably considered to be so, without the inclusion of M479 in the genotype panel.

None of the defining SNPs under R2 (including M479) look like they've been confirmed with your Kashmiri friend; the results simply suggest they're some flavour of R-M207 that isn't R1-M173 or R2a-M124.

Specific testing of M479 would help massively with deciding whether they're an actual variant of R*-M207 or R2*-M479.

[khanabadoshi]

I think my friend will be open to additional testing. Should he do y-STR test from ftdna or specific SNP test? Any suggestions?


While we are on the topic of interesting results. Another friend of mine whom I've tested has an uncommon haplotype. I believe R1b is not common at all in the region? (I'm not very well read in y-DNA halpgroups, so correct me if I am wrong). His father is a Syed and his mother a Durrani; they are from Multan.



I followed the same technique as I did with my Kashmiri friend:





y-SNP in .csv table format.
y-SNP in text document format.

[Huntergatherer1066]

I think my friend will be open to additional testing. Should he do y-STR test from ftdna or specific SNP test? Any suggestions?

He would need to do at least Y-DNA12 through FTDNA (you can order them through group project pages) if you wanted to do further SNP testing there, Y-DNA12 + M479 individual SNP test would be the cheapest next step as far as FTDNA goes I guess. The SNPs are the key but STR results would also be interesting to see, and also if by chance he had any matches.

[khanabadoshi]

He would need to do at least Y-DNA12 through FTDNA (you can order them through group project pages) if you wanted to do further SNP testing there, Y-DNA12 + M479 individual SNP test would be the cheapest next step as far as FTDNA goes I guess. The SNPs are the key but STR results would also be interesting to see, and also if by chance he had any matches.

Hopefully, I will be seeing him in a month or so. I'll surprise him with an ftdna swab haha.