Hey Sumskilz, I was looking at these two PCA plots:
Spoiler Alert, click show to read:
They're both from the same study I believe (The genome-wide structure of the Jewish people) but they're showing different results. Any idea why?
When you're creating a PCA, you're dealing with data that can be divided into more dimensions than two, but you're trying to map them onto only two dimensions. Normally you'll see just PC dimensions 1 and 2 (PC1 and PC2) because they're statistically most significant. But the result of just using PC1 and PC2 is that it obscures some of the finer scale data.
Supplementary Note 2:
We investigated the lower-ranked PC3 and PC4 for meaningful variation (Supplementary Fig. 2b and 2c) and describe each PC in terms of the two underlying genotype/allele frequency gradients, which can be inferred when SNPs loading the PC are extracted and their frequencies are estimated within population samples. When the graphs of allele frequencies are superimposed on the PC plot (data not shown), two opposing gradients are evident, each having the highest frequencies on the two extremes of the PC plot. Allele frequency gradients beginning on each side decline towards the zero point of the plot, where they overlap or disappear. We used this general methodology to describe coordinates of populations on the PC plots. Indeed, the analysis of the Old World dataset suggests that PCs beyond the top two add some extra information relevant to the subject of our study.
Without getting into all the math, the second plot is an attempt to warp and smooth the data from PC1 and PC2 by using some information from PC3 and PC4. If you tilt it counter-clockwise a bit, you can see the the second one (warped with the lower ranking PCs) starts to resemble the geographic distribution better, which is how you'd expect it to be for all but the most genetically isolated groups.
In this PDF, the first plot is the same as in the main paper (PC1 against PC2), but the second plot is PC1 against PC3, and the third plot is PC1 against PC4.
Originally Posted by Enros
You don't seem to be familiar with how the burden of proof works in when discussing social justice. It's not like science where it lies on the one making the claim. If someone claims to be oppressed, they don't have to prove it.
I found some face composites made using the faces of the athletes that competed in the 2012 London Olympics. It's kind of cool seeing the differences in phenotypes.
https://imgur.com/a/uW3mk
Last edited by Blaze86420; September 14, 2016 at 05:21 PM.
I found some face composites made using the faces of the athletes that competed in the 2012 London Olympics. It's kind of cool seeing the differences in phenotypes. https://imgur.com/a/uW3mk
Yeah, those are cool. You can tell the sample is significant when the male and female composites start to look like siblings, which I think is the case for most of them.
The Arabian male looks pretty much exactly how I'd expect. Turks only have 10-20% Central Asian ancestry, but I think you can see it in those composites' face shape a bit compared to nearby populations. To me, the Levantine ones look a lot more like Southern Europeans than most of the others do, which you'd expect based on DNA. The Israelis look really European compared to the Tel Aviv composites I have:
Spoiler Alert, click show to read:
I suspect that's because they're mostly from upper class families (all those wealthy Ashkenazi guys that want to marry blonde women), but it still surprises me because the male one looks more European than me and I'm genetically ~65% European. That is probably a selection bias that effects most of the Olympics composites, but probably doesn't make as much of a difference in more homogeneous populations.
I may as well put the other composites I have here as well:
Spoiler Alert, click show to read:
Damascus:
Ankara:
Greek Cypriots:
Turkish Cypriots:
Originally Posted by Enros
You don't seem to be familiar with how the burden of proof works in when discussing social justice. It's not like science where it lies on the one making the claim. If someone claims to be oppressed, they don't have to prove it.