tag:blogger.com,1999:blog-2845420051498655579.post6399930494794919704..comments2023-10-22T05:35:35.176-07:00Comments on Vaêdhya: Steppe Ancestry Estimations in West, Central & South Asia (Ancestral Proportions Method) [Original Work]DMXXhttp://www.blogger.com/profile/16689998564656086919noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-2845420051498655579.post-85878223548084155092015-08-09T22:00:15.302-07:002015-08-09T22:00:15.302-07:00Thanks for the comment. I fully agree with all poi...Thanks for the comment. I fully agree with all points raised. Given the inadequate aDNA coverage of the region, none of the tools or methods we currently have work as a sufficient workaround for addressing the genetic architecture issue. ADMIXTURE is certainly no exception.<br /><br />The intention here was to derive as much informativeness from ADMIXTURE as is currently possible. As you pointed out, the accuracy of the numbers are reliant on the methodology used to generate them. I suspect these numbers serve as a lower bound regarding what we'll see when Bronze Age aDNA emerges from West, Central and South Asia. Based on the uniparental data f.ex., I suspect the Pamiri Tajiks will be between 45-55% Sintashta-derived. DMXXhttps://www.blogger.com/profile/16689998564656086919noreply@blogger.comtag:blogger.com,1999:blog-2845420051498655579.post-43329701588226836332015-08-08T00:04:55.997-07:002015-08-08T00:04:55.997-07:00I'm not sure it's possible to use the Andr...I'm not sure it's possible to use the Andronovo samples to validate these results, mostly because of the accumulation of a lot of recent local drift among present-day populations.<br /><br />The recent drift is a problem because it affects the Admixture results. For instance, a lot of Sintashta-derived ancestry might be hiding in the South Asian-specific components. Sintashta and Andronovo won't show any membership in such clusters for obvious reasons, but there's no reason to say they're not partly ancestral to them.<br /><br />On the other hand, methods based on f4 stats, like qpAdm, in theory might suffer from going back too far in the phylogeny and thus ignoring a lot of informative drift. So the only way to solve this problem is to compare Bronze Age samples from West and South Central Asia to Sintashta and Andronovo. Davidskihttps://www.blogger.com/profile/04637918905430604850noreply@blogger.com