/sci/ - Science & Math

I don't see how statements like "genetic closeness" are scientifically meaningful and why these arguments are so common (e.g. "members of each race will be closer related to some members of another race than to some members of the same race"), To me those are all rather trivial statements.
I don't see how, given enough data, there can be any room for this kind of reasoning.
I would say the most unbiased view on race would be searching for clusters in high-dimensional space;
the human genome has about 20k genes. So we are talking about a 20k-dimensional space, which by itself would naturally be an unmanageable amount.
However, it will be very likely that even if we had genetic info of all people on earth the data would be a very "flat" manifold in that space, where many genes are likely to weakly differ because of their high sensitivity. These can be weeded out alongside with genes that themselves are highly correlated, reducing redundancy.
I guess the only real limitations are from sampling?