BayesFM: a software program to fine-map multiple causative variants in GWAS identified risk loci.

We herein describe a new method to fine-map GWAS-identified risk loci based on the Bayesian Least Absolute Shrinkage Selection Operator (LASSO) combined with a Monte Carlo Markov Chain (MCMC) approach, and corresponding software package (BayesFM). We characterize the performances of BayesFM using simulated data, showing that it outperforms standard forward selection both in terms of sensitivity and specificity. We apply the method to the NOD2 locus, a well-established risk locus for Crohn9s disease, in which we identify 13 putative independent signals.