South Asian HGDP Haplogroups

[MonkeyDLuffy]

No worries. We have to wait until the Punjab is properly sampled. As you can see Sindh has 6 Y7 of 11 R1a1.

Are those Y7 Lohana?

[kamil154]

mtDNA A - looks more northern.
See eg. Leh Brokpa
"They speak Dardic, a sub-branch of the Indo-European language family ... 75 random Brokpa males from the Dah-Hanu region, on the northern bank of the Indus river ... The mtDNA profile shows a predominance of mtDNA HG A4 that must have arrived from outside the Indian subcontinent." https://www.tandfonline.com/doi/abs/...nalCode=iahb20

On the Y side mainly L-M357
"the Brokpa males possessed a high proportion of NRY-HG-L1a2(M357) (62.7%)"
https://www.tandfonline.com/doi/supp...700?scroll=top

This actually makes sense. My maternal ancestry is from a Kashmiri Bakhshi family and it could mean there was a female ancestor from Ladakh at some point. A4 is also somewhat common in Xinjiang and Tibet.

[Ebizur]

This actually makes sense. My maternal ancestry is from a Kashmiri Bakhshi family and it could mean there was a female ancestor from Ladakh at some point. A4 is also somewhat common in Xinjiang and Tibet.

Haplogroup A is the third most common mtDNA haplogroup among Tibetans (after haplogroup M9 and haplogroup D), with the breakdown of the A clade being as follows according to Figure S3 of Xuebin Qi et al. (2013), "Genetic Evidence of Paleolithic Colonization and Neolithic Expansion of Modern Humans on the Tibetan Plateau":

7.87% A-16362 (this clade may be known as "A4"; the T16362C mutation is shared among nearly all members of haplogroup A except those who belong to the A10, A5, A8, A11, A3, A7, and A9 subclades)
5.58% A11(x16234)
0.95% A11+16234
0.13% A5
0.07% A7
0.03% A10

[parasar]

Are those Y7 Lohana?

Not sure, but the samples were from interior Sindh as I recall.

[Coldmountains]

Here is a spreadsheet containing maternal and paternal haplogroups for all of the South Asian HGDP cell lines I've managed to find.

Y-SNP calls were extracted from VCF data generated by Bergström et al. - "Insights into human genetic variation and population history from 929 diverse genomes" and inputted into YSEQ's Clade Finder v1.0 to determine Y-DNA subclades. Some of these HGDP cell lines weren't present in any data from this study, so their paternal haplogroups were taken from Sengupta et al. - "Polarity and Temporality of High-Resolution Y-Chromosome Distributions in India Identify Both Indigenous and Exogenous Expansions and Reveal Minor Genetic Influence of Central Asian Pastoralists". I'm still looking for WGS/chrY sequencing data for these samples but I've had no luck so far. You could probably infer their subclades from other samples in their populations though. The HGDP00111 (classed as Hazara) and HGDP00220 (classed as Pathan) cell lines originate from the same individual, who is most likely Hazara judging from the Y-DNA subclade and autosomal results posted elsewhere.

mtDNA subclades were determined by downloading FASTA files from these cell lines submitted to GenBank by Lippold et al. - "Human paternal and maternal demographic histories: insights from high-resolution Y chromosome and mtDNA sequences" and running them through HaploGrep v2.1.25 to determine mtDNA subclades.

If there are any errors then please let me know.

Link: https://docs.google.com/spreadsheets...#gid=799017964

Some of the samples were uploaded on Yfull already but many not. Is there some way to ask Yfull to upload the remaining samples? Or is the sample quality not good enough? I would really really like to see the remaining Y6 samples on Yfull to finally get more clades and snps for Y6 defined. Else we Y6 and other Central/South Asian Y-Dna carriers need to wait very long for more private samples being uploaded there.

[aaronbee2010]

Some of the samples were uploaded on Yfull already but many not. Is there some way to ask Yfull to upload the remaining samples? Or is the sample quality not good enough? I would really really like to see the remaining Y6 samples on Yfull to finally get more clades and snps for Y6 defined. Else we Y6 and other Central/South Asian Y-Dna carriers need to wait very long for more private samples being uploaded there.

Most of those files were from Bergstrom et al. 2020, which sequenced (next-generation WGS) their HGDP samples to around 30-40x coverage, so the quality is definitely ideal, despite capping the MAPQ of aligned reads at 20 for some reason (they used BWA-MEM for FASTQ alignment, which normally has a MAPQ cap of 60).

I see a lot of those HGDP samples coming under R1a-Y920*, which I would definitely like to see resolved further as these samples might share additional SNPs with each other that are not currently on the YTree (or other trees).

It may be worth having a look to see if any other HGDP samples have been uploaded to YFull since I made that spreadsheet but I don't think there will be much. I believe YFull would be more inclined to upload another sample (after their initial round of HGDP uploads which seems to have concluded now) if they knew it would cause some sort of addition to the YTree.

If you like, I can take a look at those samples and see if I can identify new SNPs in them for you, as the presence of such SNPs could be an incentive for YFull to add those samples. If I do, I can submit those SNPs to YBrowse, as I've done this before for some R2 samples I've analysed previously.

[Coldmountains]

Most of those files were from Bergstrom et al. 2020, which sequenced (next-generation WGS) their HGDP samples to around 30-40x coverage, so the quality is definitely ideal, despite capping the MAPQ of aligned reads at 20 for some reason (they used BWA-MEM for FASTQ alignment, which normally has a MAPQ cap of 60).

I see a lot of those HGDP samples coming under R1a-Y920*, which I would definitely like to see resolved further as these samples might share additional SNPs with each other that are not currently on the YTree (or other trees).

It may be worth having a look to see if any other HGDP samples have been uploaded to YFull since I made that spreadsheet but I don't think there will be much. I believe YFull would be more inclined to upload another sample (after their initial round of HGDP uploads which seems to have concluded now) if they knew it would cause some sort of addition to the YTree.

If you like, I can take a look at those samples and see if I can identify new SNPs in them for you, as the presence of such SNPs could be an incentive for YFull to add those samples. If I do, I can submit those SNPs to YBrowse, as I've done this before for some R2 samples I've analysed previously.

Thanks for your answer. Good to know that the quality is good so there is some hope for Y920* guys like me. I believe my Pashtun Y6 should be close to these samples, but there is also 10% L657 among Tajiks and Uzbeks in the North, which subclades are unknown so far. Based on Pashtuns orginating in Arachosia my clade should be closer to these Gedrosian Y920* than to L657 Tajiks/Uzbeks in the North (some of it could be recently from Pashtuns like some of the deep tested Z2123) but this can not be confirmed yet.

It would be really great if you could identify new SNPs. I am probably soon going to buy a WGS test but knowing that there are almost no samples on Yfull from the region and being Y920* i felt like this would be pretty disappointing for some years before more samples are uploaded. At least i see now among many Afghans interest in genetic testing so we will likely see a big number of new Afghan samples on Yfull in the next years. Some future study focusing on L657 like the Arpad study focusing on Z2123 would be a dream.

[Jatt1]

R1a and J2 are frustrating to have since ultimate origins can be a bit ambiguous.

I was hoping for either something more clear like J1/E on one side or H/L on the other side.

What do you mean ultimate origins, how is it more clear for J1/E?

[ssamlal]

I'm a possible Y-DNA match to HGDP00254

He appears to be Pathan from Pakistan based on this: https://www.sciencedirect.com/scienc...02929707623532

Attachment 39151

YFull recently added 2 downstream branches and moved HGDP00254 to H-Z5867 > Y61050*. The Punjabi Hindu and I moved to H-Z5867 > Y68923

YFull H-Y68923.png

[aaronbee2010]

Thanks for your answer. Good to know that the quality is good so there is some hope for Y920* guys like me. I believe my Pashtun Y6 should be close to these samples, but there is also 10% L657 among Tajiks and Uzbeks in the North, which subclades are unknown so far. Based on Pashtuns orginating in Arachosia my clade should be closer to these Gedrosian Y920* than to L657 Tajiks/Uzbeks in the North (some of it could be recently from Pashtuns like some of the deep tested Z2123) but this can not be confirmed yet.

It would be really great if you could identify new SNPs. I am probably soon going to buy a WGS test but knowing that there are almost no samples on Yfull from the region and being Y920* i felt like this would be pretty disappointing for some years before more samples are uploaded. At least i see now among many Afghans interest in genetic testing so we will likely see a big number of new Afghan samples on Yfull in the next years. Some future study focusing on L657 like the Arpad study focusing on Z2123 would be a dream.

So I downloaded the 4 HGDP R1a-Y920* samples as well as 3 control samples who come under other subclades of R1a-Y920 and performed variant calling on all 7 BAM files simultaneously and I found two distinct subclades, each harbouring two of those R1a-Y920* HGDP samples.

When I compared the SNPs in those subclades to YBrowse's VCF file, I noticed that every single SNP in both subclades was already on the database, with the majority of those SNPs harbouring the "FT" prefix, which means those SNPs were discovered by FTDNA already (guess they beat me to it this time lol). The only problem is that FTDNA do not disclose the public samples on their Haplotree under a given subclade, at least from what I've seen.

This makes sense as FTDNA have recently started analysing 1000Genomes and HGDP samples to further refine their Haplotree (although it appears that private FTDNA testers are also present under these two subclades). I then searched said Haplotree and funnily enough, I ended up finding two subclades of R1a-Y920 not on YFull's tree, whose SNPs corresponded almost exactly to the two subclades I found. Here's the structure of these two subclades with FTDNA nomenclature:

R1a-Y920 > SK2051 (31 SNPs)
R1a-Y920 > SK2051 > FT281076 (4 SNPs)
R1a-Y920 > Y148366 (1 SNP)
R1a-Y920 > Y148366 > Y149047 (4 SNPs)
R1a-Y920 > Y148366 > BY181409 (8 SNPs)
R1a-Y920 > Y148366 > BY181409 > FT281004 (13 SNPs)

Based on these findings, here are the FTDNA haplogroup assignments of the 4 HGDP samples that would currently be designated as R1a-Y920* on YFull:

HGDP00076 (Balochi): R1a-Y920 > Y148366 > BY181409 > FT281004
HGDP00096 (Balochi): R1a-Y920 > SK2051 > FT281076
HGDP00150 (Makrani): R1a-Y920 > Y148366 > BY181409 > FT281004
HGDP00199 (Sindhi): R1a-Y920 > SK2051 > FT281076

Unfortunately, the four HGDP samples on their own can only generate two additional subclades on YFull, not the six on FTDNA, as the additional private samples responsible for this split don't appear to be on YFull at the moment, so bear that in mind when choosing between a Big Y-700 and a WGS. I'll see if I can resolve things further from the other public samples on YFull, but no promises there.

I believe Nebula and FTDNA were planning a partnership where you pay a fee for FTDNA to interpret your raw data and determine haplogroups + matching (assuming discovery of private SNPs is here too) but progress on this seems to have remained frozen for quite some time.