Dan MacArthur at Genetic Future has the details. Some of the stuff coming out of genomics reminds me a lot of what you see with social science; lots of sexy studies which turn out not to be as significant upon later analysis. Perhaps hypotheses are overrated?
Perhaps hypotheses are overrated?
Well, some of the people doing genome scans would argue that they're going in without any hypothesis at all.
Of course, the hypothesis they're implicitly testing is that common diseases are caused by common variants that can be tagged with a SNP chip; as it turns out, this hypothesis is largely wrong.
Of course, the hypothesis they're implicitly testing is that common diseases are caused by common variants that can be tagged with a SNP chip; as it turns out, this hypothesis is largely wrong
How do you know they are wrong?