The authors of a controversial genetics paper published last year in Science published a retraction today , acknowledging "technical errors" in their gene-finding strategy. The work, led by Paola Sebastiani and Thomas Perls of Boston University, claimed to have found a "signature" of 150 gene variants  that together could help predict whether someone might live to be 100.
But within days of the paper's appearance, critics charged that the authors hadn't accounted for intrinsic flaws in gene microarrays they used to find those variants, among other problems. Last July, Perls and Sebastiani told Science  that they were unaware of problems in the microarrays and were working hard to verify their results.
In today's retraction, the authors do not reveal how severely their original findings were affected by the errors. "We feel the main scientific findings remain supported," they write, but "the specific details of the new analysis change substantially from those originally published." In an e-mail exchange, Perls wrote that the researchers "are very anxious to get our corrected results out into the scientific literature," and until then, can't say more about what they are.