Last week, BioNews reported on a study published in Science that claimed to have identified several gene clusters associated with longevity. The study drew significant media interest but, following the paper's publication, experts have raised concerns about the data.
Commenting in Newsweek, both Dr Kari Stefansson (founder of deCODE Genetics) and Professor David Goldstein (Duke University) highlighted potential flaws in the data.
Dr Stefansson raised concerns that because the study used different DNA scanning systems for the centenarians and controls, this could lead to 'genotyping bias'. He was especially worried about one of the 'gene chips' used, the 610-Quad manufactured by Illumina, because this chip has been shown to have a number of quirks that may account for the associations seen.
Professor Goldstein also highlighted how these types of analyses can produce experimental differences that impact on the results produced, if different chips are used or if analyses are carried out in other laboratories. Other researchers were concerned that the sample size of 1,000 centenarians and 1,200 controls was too small for this type of genome-wide association study (GWAS).
But the study authors Dr Thomas Perls and Professor Paola Sebastiani from Boston University were reported to be 'standing by their findings'. Dr Perls, for example, was quoted saying: 'there have been very premature and harmful proclamations about the data' and that 'good science takes time' (Science Insider, 8th July).
The discovery of the apparent 'flaws' in this data also raise wider questions about the validity of such studies and why the journal Science did not pick up on these problems sooner, Discover Magazine's Gene Expression blog said.