Building Biomedical Infrastructure for Aggregating Variant Annotations and Prioritizing Disease Variants

Includes bibliographical references (pages 25-27).The recent advent of massively parallel sequencing technologies has allowed academic and clinical research to shift focus into genome wide discovery of variations for investigation of their role in disease. Attempts to comprehensively annotate genetic variants for population frequency, pathogenicity, and clinical relevance are distributed across numerous efforts. Currently, annotating genomes and variants involves downloading and parsing flat-files and uploading the data into a database. This becomes especially tedious when studying complex and heterogeneous diseases with respect to variant type or when annotations need to be updated. In order to alleviate fragmentation and allow systematic interpretation of genomic variation data, we have developed MyVariant.info.\ud MyVariant.info is an aggregation of human genetic variant databases into a unified, queryable web service. Development required aggregating and loading data, developing a programmatic interface for interactively querying and exploring annotations, and integrating the service into a variant discovery pipeline. I will also demonstrate a usage case for prioritizing variants from whole exome studies for their role in disease