I'm not an expert, but the idea is that mutation rates indicate age. If a bunch of people have a particular pattern, but they're all a little bit different to t

I'm not an expert, but the idea is that mutation rates indicate age. If a bunch of people have a particular pattern, but they're all a little bit different to the same degree of difference, you can figure out how old that pattern is. Presumably the changing bits must be in junk DNA otherwise mutation rates would be affected by selection pressure.