PDA

View Full Version : An old one, but a goodie: PCA depicting insignificant variability of OOA populations



Eugene Finkelstein
03-28-2023, 12:32 PM
Just thought to share it here. Old work. Complex Patterns of Genomic Admixture within Southern Africa (https://journals.plos.org/plosgenetics/article/figure?id=10.1371/journal.pgen.1003309.g002), PLOS Genetics 9(3): e1003309.
But what's precious in it is that the authors don't calibrate carefully their PCA so the East Asian (ASEAN Plus Three) and West Eurasian (Europe, Middle East, & Southern Asia) populations shown as the 2 equal counterpats to the African grouping, distant from each other samely as from sub-Saharan Africa & variative inside samely as Africans variative inside. This old work not intended to do so, so, the OOA populations shown as they are - everybody outside of SSA are pretty the same. Difference between Spaniard macho from Andalusia and Native American chief from Great Plains is tiny if compared even to the two Negroes from some Tanzanian village, and if compare macho & chief not to the predominant Negroe populations of Africa, but to the Khoe San HG, well, then the difference not even tiny, but barely visible at all.
https://i.postimg.cc/vbW5NV5y/pgen-1003309-g002.png

Eugene Finkelstein
03-28-2023, 12:36 PM
By the way, do anybody of you folks possesses Whole-genome sequencing reveals a complex African population demographic history and signatures of local adaptation (https://www.cell.com/cell/pdf/S0092-8674(23)00101-0.pdf) article, published this year - very new, but unfortunately paywalled and Sci-Hub & LibGen currently don't have it... If somebody do - please, share, I'll greatly appreciate. It looks like they've made pretty good new human populations tree, would be happy to see. So, if anybody can share, it would be really cool, folks.

Petalpusher
03-28-2023, 12:50 PM
Tends to do that with a K2 using an ENA component which is everything non African. It's usually more for archaics or very ancient Eurasian genomes, Nenderthal/Denisovians, Paleolithics or indeed like it's the case here for specifically African sub structures.

Even with a regular PCA and a lot more components, the variation of all Europeans is within the same ball park as say the Chineses and the Koreans.

At larger scale and while the difference is not large obviously, the furthest from Chimps today are Europeans (TSI and CEU here), the closest Africans, and East Asians (+ NAs) are in between. Possibly only because Euro are more exclusively Neanderthal admixed.

https://www.researchgate.net/profile/Federico-Sanchez-Quinto/publication/232321487/figure/fig2/AS:300587263512577@1448676996281/PCA-analysis-of-North-African-Sub-Saharan-European-and-Asian-populations-Upper-right.png

Eugene Finkelstein
06-05-2023, 02:58 PM
By the way, do anybody of you folks possesses Whole-genome sequencing reveals a complex African population demographic history and signatures of local adaptation (https://www.cell.com/cell/pdf/S0092-8674(23)00101-0.pdf) article, published this year - very new, but unfortunately paywalled and Sci-Hub & LibGen currently don't have it... If somebody do - please, share, I'll greatly appreciate. It looks like they've made pretty good new human populations tree, would be happy to see. So, if anybody can share, it would be really cool, folks.

https://gwern.net/doc/genetics/selection/natural/human/2023-fan.pdf