Scaling up Genetic Analysis

ABSTRACT Sequencing technology determines the need for genome analysis tools that meet the challenges of scale. I will describe our efforts to develop a software package, hail, that uses spark and scala in a distributed model of computing to achieve large scale genome analysis and quality control. We can perform primary quality control analyses on whole genome sequencing datasets of ~5,000 individuals in under an hour. Using hail, we have performed analyses of education attainment on a sample of over 14,000 individuals, identifying a clear role of ultra-rare disruptive mutations. We further explored this class of variation across a wide range of traits and demonstrate that neuropsychiatric traits appear to have a directional burden effect in contrast to later onset systemic disease.