Abstract
The results from association studies are usually summarized by a measure of evidence of association (frequentist or Bayesian probability values) that does not directly reflect the impact of the detected signals on familial aggregation. This article investigates the possible advantage of a two-dimensional representation of genetic association in order to identify polymorphisms relevant to disease: a measure of evidence of association (the Bayes factor, BF) combined with the estimated contribution to familiality (the attributable sibling relative risk, lambdas). Simulation and data from the North American Rheumatoid Consortium (NARAC) were used to assess the possible benefit under several scenarios. Simulation indicated that the allele frequencies to reach the maximum BF and the maximum attributable lambdas diverged as the size of the genetic effect increased. The representation of BF versus attributable lambdas for selected regions of NARAC data revealed that SNPs involved in replicated associations clearly departed from the bulk of SNPs in these regions. In the 12 investigated regions, and particularly in the low-recombination major histocompatibility region, the ranking of SNPs according to BF differed from the ranking of SNPs according to attributable lambdas. The present results should be generalized using more extensive simulations and additional real data, but they suggest that a characterization of genetic association by both BF and attributable lambdas may result in an improved ranking of variants for further biological analyses.