Genome wide association studies (GWAS) have become a routine tool to analyze the genetic basis of complex traits. Thousands of genomic loci have been implicated for a variety of phenotypes using this design and a fairly simple statistical analysis.

There is evidence, however, that a number of relevant associations are yet to be discovered. Within this framework, we have worked to outline analysis strategies that might improve power, while controlling the FDR of the discovery of genetic loci with phenotypic effects. I will summarize our progress in two directions: in one case we capitalize on the multiplicity of available phenotypes, in another we adapt to the multivariate nature of the genetic signature. While many contributed to this work, let me single out C. Peterson, Y. Benjamini and M. Bogdan.