[eblume]

I used to work for a biotech company that shared a somewhat similar field to 23andme, although aimed at prenatal diagnostics. We were doing the exact thing that #3 on their list claims. While I am not a scientist, my understanding is that the 23andme has a reputation for being very sloppy at science. It's perhaps useful as a source of personal information but I wouldn't trust anything they talk about, as far as research goes.

Point in case, the whole 'noninvasive prenatal diagnostics' thing was something we had down to a commercial clinical procedure... in 2011. In fact we had it down pretty good in 2009 (before I was at the company), but it takes a long time to get through certification with these sorts of tests. My understanding is that research groups had been doing that sort of thing since 2004 - I don't know when the first human fetus was sequenced but it was certainly not last October.