Variant interpretation in the genomic diagnosis of rare disease is currently a highly manual
process, often requiring hours of labour by skilled curators; this means the process is not scalable, and also inhibits the reanalysis of previously generated data - meaning that
families that cannot be diagnosed during their first clinical analysis may miss out on a
diagnosis entirely. In this project, we are collaborating with the Broad Institute and Microsoft to combine very large, stringently curated data sets from a variety of sources with machine learning approaches run on Azure to create robust tools for prioritising the variants found in patient genomes, dramatically speeding up the curation process. These methods will be tested on research data spanning tens of thousands of families, and then deployed on real-world clinical data from diagnostic labs in the US and Australia. Find out more here.