There are now several confirmed common variants that influence common diseases, responses to infection, and responses to drugs. For most common diseases however all common variants identified explain only a few percent of the known heritability, and many of the signals emerging from genome wide association studies have yet to be tracked to single common variants, raising the possibility that in some cases the signals emerge from associated sets of rare variants. Here I argue that progress in identifying the so called ‘missing heritability’ for many human traits will be facilitated using an extreme phenotype whole-genome sequencing paradigm. I illustrate the basic structure and motivation for this approach using examples from our work on host determinants of control of HIV-1.
