There have been sparce tests of genome-wide associations between gene expression and longevity in humans, but more are coming up the pipeline. Kerber et al (2009) used a dataset of Utah grandmothers to test 2151 always expressed genes in a proportional hazards model. The predicted mortality from their six gene model was able to account for 23% of the variance in predicted mortality. Here are the genes in their multivariate model as well as the coefficients (which are rough indicators of how important the gene is to the model):
- CORO1A (-0.27),
- FXR2 (0.21),
- CBX5 (-0.074),
- PIK3CA (-0.0094),
- AKAP2 (-0.0086),
- CUL3 (-0.0081).
A very cool study and hopefully there will be more like it soon.
Kerber RA, et al. 2009 Gene Expression Profiles Associated with Aging and Mortality in Humans DOI: 10.1111/j.1474-9726.2009.0046.