Lies, Damned Lies And …

Thanks to new tools like the false-discovery-rate technique, they've repeatedly shot down claims. In searching for genes that raise your risk of developing some disease, for instance, geneticists applied a traditional test: is a link to a gene less than 5 percent likely to be a coincidence? Unfortunately, when you're testing thousands of genes, that lets too many false positives through. "Most of the literature [linking a gene to a disease] is riddled with false discoveries," warns Fred Wright of the University of North Carolina. Something to consider before forking over thousands of dollars to have your DNA analyzed by companies marketing the tests to consumers.

Geneticists now acknowledge that scores of claims about disease and DNA are wrong, and have tightened their criteria for what's likely to be a real link. Unfortunately, "the process of applying a stricter threshold [to keep out false links] leads you to overestimate" what proportion of the risk of getting a disease is due to DNA rather than, say, lifestyle, says Wright. At next month's meeting, he and UNC colleagues will unveil a new statistical tool "to tamp down that inflation and get more realistic estimates of the magnitude of a gene's effect." In applying it to some highly publicized studies, such as one linking genes to diabetes, he has knocked down genes' importance by as much as one third. If Disraeli were alive, he might amend his line to "lies, damned lies, revealed by statistics."

© 2008