S/SC ASIAN Results :: 23andMe

[drobbah]

Is the trace Sub-Saharan African I find in a lot of Gujarati Muslim relatives indicating something real? I also find trace Southern European in some cases, which I used to score on ancestry. Is this common as noise?


Attachment 42662

Ethio/Eritrean & Somali percentages probably came indirectly from Peninsular Arabs.Some of it might also be real direct ancestry as the Muslim sultanates of the Horn had strong contacts with Gujarat.In the Futuh Al Habasha, a book that details the 14th to 15th century wars between Abyssinia and the muslim sultanates indicated that many of the Christian nobility who were captured in battle or raids would be sold in to slavery with many of these ending up in Gujurat.One of the oldest copies of this book was also sent to Gujurat where it was preserved.The most famous Horner slave was Malik Ambar who defeated the Mughals and ruled a Deccan kingdom, he hailed from a now extinct ethnic group.

Here's a source explaining the history of it:
futuhgujurat.jpg

[subzero85]

Is the trace Sub-Saharan African I find in a lot of Gujarati Muslim relatives indicating something real? I also find trace Southern European in some cases, which I used to score on ancestry. Is this common as noise?

Attachment 42665
Attachment 42663
Attachment 42661
Attachment 42664
Attachment 42662

Not sure about your exact community.

But Dawoodi Bohra's were found to contain some African mtDNAs in this old study: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2859343/

My pet theory for their L maternals is Eritrea/Ethiopia/Somalia => Yemen => Gujarat

[drobbah]

Not sure about your exact community.

But Dawoodi Bohra's were found to contain some African mtDNAs in this old study: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2859343/

My pet theory for their L maternals is Eritrea/Ethiopia/Somalia => Yemen => Gujarat

L0a2a2 mentioned in the paper isn't exactly a Horner maternal lineage.Probably it came directly to India like the Siddis or with an admixed Yemeni, Hadhramawt for example has extremely high frequencies of maternal Bantu lineages (25%) and autosomal ancestry

[MuslimPatel123]

As great as IBD matching is, they are quite obviously using PCA projections akin to g25 only they do a worse job by far. After extensive examination of 23andme's African components, this is clearly what's going on - this is compounded by the fact the reference pool is relatively shallow and can't be bolstered by public sample sets precisely because of lack of overlap like you said. The imbalance is still unlikely to be rectified as long as they are quantifying ancestry via PCA projection without a deep enough understanding of the reference groups at their disposal. The recent paper, although only concerning African slave descendants, sheds lights on 23andme's ineptitude - if they employ IBD segment matching so well in recent ancestry computation as you believe, why do the results of that paper contradict their commercial autosomal presentations? What makes you think this isn't happening with all non-European references? Guaranteed South-Asians are being treated in the same way. Admixture artefact simulation. This isn't an issue of the reference pool, but behind the stats themselves. Not harsh at all, their methods are trash.

...

I'm getting 0.8% "Malayali Subgroup" in my latest results. The description for the community is:
"This genetic signature reaches high levels among Christian communities who live in Kerala, a large number of whom immigrated to the United States during the last century."
Looks like it might be caused by the fact that some Kerala Christians also have a middle eastern shift, resulting in the naive 23andme algorithm to error out on my segment.

[ndubbpk]

They score similar on Harappa to a Khetran Baloch result I've seen previously. khana posted a Seraiki Baloch result before and they scored more or less exactly like the academic HGDP Baloch samples. Though, to be specific, the Seraiki Baloch he posted is actually from from central Sindh but settled for many generations. They are Sindhi speaking and unable to speak Balochi but he refers to himself as Seraiki speaking as most of his tribe and other Baloch from his hometown speak Seraiki.

Hey man kinda off topic but is it possible you could clear up some of your inbox so I can message you, please?

[Sapporo]

Hey man kinda off topic but is it possible you could clear up some of your inbox so I can message you, please?

I deleted a few messages.

[Sapporo]

Does Afghanistan only mean the Pashtun ethnicity on 23andme? Since there are other ethnicities in Afghanistan too or does it leave that ambiguous?

It does not specify. Only cities. For KPK and other Pakistani provinces, I don't believe it even specifies the city.

However, it is likely safe to assume that Kabul has a mix of samples from different Afghan ethnic groups while a city like Kandahar is predominately Southern Pashtun samples or Nangarhar is mostly Northeastern Pashtun samples. Tajik dominated cities or provinces (such as Parwan or Panjsher) are primarily Tajik, etc. A city like Ghazni probably has a mix of Tajik, Pashtun and Hazara samples.

[subzero85]

https://www.reddit.com/r/23andme/com..._of_pakistani/

Kashmiri Pakistani results

Paternal: H-M2914
Maternal Haplogroup: H13a2a

[tipirneni]

https://www.reddit.com/r/23andme/com..._of_pakistani/

Kashmiri Pakistani results

Paternal: H-M2914
Maternal Haplogroup: H13a2a

H-M2914 is a major clade among Kammas in Andhra Pradesh

[AlluGobi]

This guy's one of my matches (very distant), family name says Syed.

23andmesyed.JPG


Paternal: R-Y7
Maternal: M4b2