What ancient component forms the bulk or majority for typical South Asian genetics?

[misanthropy]

Also feel free to go by region (e.g. NW India, NE India, etc.). I have a feeling people are going to say IVC, but then again that has its components too.

[poi]

Although the Harvard/ReichLab Narasimhan study did omit a bulk of NW populations for various reasons, here is a PCA based on the study's supplement datasheet. They broke up the components into 3 parts: AASI-related, Indus-diaspora-related, and Steppe-MLBA-related.

https://www.biorxiv.org/content/earl...1.figures-only

Biplot
https://i.imgur.com/rITreLb.png


PCA - Variance coverage is ~100%
https://i.imgur.com/2PaZb8A.png

[MonkeyDLuffy]

Although the Harvard/ReichLab Narasimhan study did omit a bulk of NW populations for various reasons, here is a PCA based on the study's supplement datasheet. They broke up the components into 3 parts: AASI-related, Indus-diaspora-related, and Steppe-MLBA-related.

https://www.biorxiv.org/content/earl...1.figures-only

Biplot
https://i.imgur.com/rITreLb.png


PCA - Variance coverage is ~100%
https://i.imgur.com/2PaZb8A.png

The study was too sketchy, especially the labeling of populations.

[redifflal]

That study's proposed model's tenacity in European equivalent would be the following hypothetical scenario:

We previously had aDNA of Balangoda man and Ust-ishim. Now we have finally found aDNA retrieved from Neolithic Anatolia and modeled modern European populations in a three way mix. We show that Anatolia_N itself has no relation with Ust-ishim but has some continuity from Balangoda man, with excess component that cannot be found in Balangoda or Ust-ishim, which we will call AAWE (Ancient Ancestral Western European). We show that modern European populations can be well modeled by Anatolia_N farmers, Ust-ishim warriors and AAWE hunter gatherers.

[poi]

The study was too sketchy, especially the labeling of populations.

Could you elaborate please? I ask this because this Reich lab study is the only available study using S/C Asian aDNA. Literally the first and the only as of this morning. I guess more aDNA from S/C Asia will cause revisions etc. But what would the aDNA based alternative be? I am honestly asking and not trying to be passive aggressive lol.

That study's proposed model's tenacity in European equivalent would be the following hypothetical scenario:

We previously had aDNA of Balangoda man and Ust-ishim. Now we have finally found aDNA retrieved from Neolithic Anatolia and modeled modern European populations in a three way mix. We show that Anatolia_N itself has no relation with Ust-ishim but has some continuity from Balangoda man, with excess component that cannot be found in Balangoda or Ust-ishim, which we will call AAWE (Ancient Ancestral Western European). We show that modern European populations can be well modeled by Anatolia_N farmers, Ust-ishim warriors and AAWE hunter gatherers.

LOL! Okay! Point taken. What is your take on the South Asian components?

[MonkeyDLuffy]

Could you elaborate please? I ask this because this Reich lab study is the only available study using S/C Asian aDNA. Literally the first and the only as of this morning. I guess more aDNA from S/C Asia will cause revisions etc. But what would the aDNA based alternative be? I am honestly asking and not trying to be passive aggressive lol.



LOL! Okay! Point taken. What is your take on the South Asian components?

The Jatt Sikh samples in it, which seemed more like PJL samples. The only information that was provided that they were collected from Bathinda, that's it. It's very vague that they collected samples without checking the clans name or background or if they're legit from that group. The plot of Jatt sikhs on the graph kind of gives you an idea that they're out of place.

Second they messed up halogroup of Malta boy and another ancient which caused controversy in our forums European section with some bans and outrage.

We tried to contact the team on Twitter multiple times, and other users asked same questions, but they simply ignored it.

[poi]

The Jatt Sikh samples in it, which seemed more like PJL samples. The only information that was provided that they were collected from Bathinda, that's it. It's very vague that they collected samples without checking the clans name or background or if they're legit from that group. The plot of Jatt sikhs on the graph kind of gives you an idea that they're out of place.

Second they messed up halogroup of Malta boy and another ancient which caused controversy in our forums European section with some bans and outrage.

We tried to contact the team on Twitter multiple times, and other users asked same questions, but they simply ignored it.

Thank you for the thorough reply. If they are indeed ignoring calls for corrections, assuming they got so many things wrong, it would seem to me that their credibility would take a hit. But if the academia works this way (preprint still needing peer review), then they would let the normal (albeit slow) process play itself out. Also, is there a chance that they are correct about those haplogroups, including Maalta Boy's?

Btw, I understand about the sampling issue wrt Jatt Sikhs. But, again, how do we know about the positioning of them when the field is so new... we cant treat Gedmatch calcs as the baseline.

Note that I'm not being confrontational, just think of me playing the devil's advocate. Basically, you can't call the Reich Lab geneticists idiots, so either they must be accurate or they are sloppy as hell.

[bol_nat]

Although the Harvard/ReichLab Narasimhan study did omit a bulk of NW populations for various reasons, here is a PCA based on the study's supplement datasheet. They broke up the components into 3 parts: AASI-related, Indus-diaspora-related, and Steppe-MLBA-related.

https://www.biorxiv.org/content/earl...1.figures-only

Biplot
https://i.imgur.com/rITreLb.png


PCA - Variance coverage is ~100%
https://i.imgur.com/2PaZb8A.png

Its possible to get AASI, Indus-diaspora, and Steppe-MLBA scores of Reich with our own kits?

[MonkeyDLuffy]

Thank you for the thorough reply. If they are indeed ignoring calls for corrections, assuming they got so many things wrong, it would seem to me that their credibility would take a hit. But if the academia works this way (preprint still needing peer review), then they would let the normal (albeit slow) process play itself out. Also, is there a chance that they are correct about those haplogroups, including Maalta Boy's?

Btw, I understand about the sampling issue wrt Jatt Sikhs. But, again, how do we know about the positioning of them when the field is so new... we cant treat Gedmatch calcs as the baseline.

Note that I'm not being confrontational, just think of me playing the devil's advocate. Basically, you can't call the Reich Lab geneticists idiots, so either they must be accurate or they are sloppy as hell.

Of course we can't base opinions on gedmatch calcs, but then we've runs on different programs by Kurd, or davidski g25. It was quite sketchy the way they discribed the origin of samples. Not to mention they also pulled the paper back because of so many issues. And no, the Malta boy halogroup was corrected in revision.

Also to mention the samples were taken from a hospital in Bathinda, you can't get more vague than this. I understand you're trying to look at both sides, which we always need to.

[poi]

Its possible to get AASI, Indus-diaspora, and Steppe-MLBA scores of Reich with our own kits?

Possible yes. Considering they are conducting experiments and sending the paper for peer review, I am sure other people can replicate it. The actual execution will depend on resources and funding. These folks are backed by endowments and millions in funding. Still, smart people can probably read the paper's methodology and try things out. Way beyond my iq level though lol