In Nature, Pauline Ng, J. Craig Venter, Sarah Murray and Samuel Levy give nine recommendations to improve direct-to-consumer genetic tests. Prompted by debate on the usefulness of such testing, the team compared results from two companies that offer these tests - 23andMe and Navigenics - on 13 diseases for 5 individuals. "Despite this limited data set we find potential implications for personalized medicine," they write.
They found the raw DNA data to be accurate, but they spotted inconsistencies between the two companies' disease-risk predictions. For four diseases, the relative risk predictions completely agreed for all individuals. For seven diseases, only 50% or less of the predictions agreed between the two companies across the individuals.
Among other things, the authors urge companies to report how much of the genetic contribution of a disease can be attributed to the markers used in their test, and how much is still unknown. Companies should also focus on better communicating high-risk predictions, test for as many drug response markers as possible, and agree on the strong-effect disease markers.
Meanwhile, the research community should study markers in all ethnicities and look at how tests affect lifestyle and behaviour. The group also calls for more large, long-term studies to measure the predictive value of known markers, and more sequencing of individual genomes.
Heather Kowalski (Kowalski Communications, Washington DC, USA)
(C) Nature press release.
Message posted by: Trevor M. D'Souza