Comments on: One girl, one exome

By: Razib Khan

Razib Khan — Mon, 22 Oct 2012 16:48:02 +0000

#2, in this case i obviously wasn’t talking about GWAS. i think one of the ‘low hanging’ near term fruit with mass adoption of 100x coverage may be a better understanding of individual mutational load. re: GWAS, etc., the sample sizes are way too small. see openSNP for testament to that. but they won’t be (and 23andMe is trying to really scale this up). platform matters, but so does cost so that the data pipeline gets fat. talk to me in 2-3 years. there are firms working on this sort of thing….

By: zach

zach — Mon, 22 Oct 2012 15:53:40 +0000

the problem I see with this and the DIY approach is that the scientist’s family isn’t going to be a big enough n for most association metrics and, for many applications, the covariation due to relatedness will possibly obscure learning, not enhance it.

What are your thoughts on this: do DIY geneticists have the kind of data available to get the most of their statistics?

By: John Emerson

John Emerson — Mon, 22 Oct 2012 13:47:13 +0000

I don’t think open and DIY science is going to be the whole future, but it is going to be a substantial part of the future.

I hope that this is true for all fields. The academic monopoly has had a number of harmful effects in many areas.