There's a lot of misunderstanding about all this
If statistical analysis demonstrates that people with gene 123plusalpha are 10% more likely to be great coders than the average member of the population then that tells you absolutely nothing useful about whether you should hire Alan who has gene 123plusalpha, or Bill who has not. But it does tell you that if gene 123plusalpha is more prevalent amongst your coders then your hiring policy is effective.
It also doesn't mean that your best coders will have gene 123plusalpha. Statistics tell you nothing about that either. So a company that operated a completely effective hiring policy would have no need to find out whether an individual has that gene before they hire them, because it tells them nothing useful about that individual. However if they did an analysis they would expect to find that gene over represented in their coding staff, and that would be the result of competent hiring, not discrimination.
This is the crucial mistake that people make time after time. Statistical trends are not applicable to individuals, and a company would be enormously foolish to hire and fire on them. But they can cast light on how well your hiring and firing is functioning.