Last summer, a study published in the journal Science on the genetics of human longevity made a splash when the authors claimed that they had identified a genetic signature that could predict with 77% accuracy whether someone would live past 100. The bold findings immediately came under fire from other geneticists and scientists who read the paper and found the study’s methodological rigor lacking. One of 23andMe’s own researchers David Hinds published a detailed critique on this blog soon after, accompanied by analyses done using 23andMe’s own database.
After more than a year, the authors have formally retracted the longevity study from the scientific record. An independent lab had redone the experiment and the analyses to eliminate the major concerns brought up by the study’s critics, but Science decided that the resubmitted work no longer met its standards for genome-wide association studies.