Two studies published today demonstrate what was immediately evident from genome-wide association studies of many common diseases: the genetic variants identified account for only a small fraction of risk.

In these cases, the authors try to predict whether an individual will get type II diabetes from a number of clinical variables, as well as recently identified genetic risk factors. The genetic factors only marginally improve the prediction of diabetes, likely to a clinically insignificant extent.

This was obvious, of course, from the initial studies themselves--you can't expect variants responsible for a meager fraction of overall disease risk to function as effective predictors of the disease. But somewhat notable nonetheless.

Labels: disease, Genetics