PGRNseq: a new sequencing-based platform for high-throughput pharmacogenomic implementation and discovery. A. S. Gordon1, J. D. Smith1, Q. Xiang2, M. L. Metzker2, R. A. Gibbs2, E. R. Mardis3, D. A. Nickerson1, R. S. Fulton3, S. E. Scherer2 1) Genome Sciences, University of Washington, Seattle, WA; 2) Human Genome Sequencing Center, Baylor College of Medicine, Houston, TX; 3) Washington University Genome Center, St. Louis, MO.
Understanding the genetic basis of an individuals response to therapeutic drugs (pharmacogenetics) is a unique area of research with significant translational impact for medicine. Although pharmacogenetics has a rich history, we lack a complete picture of the common and rare genetic variation that influences an individuals response to medication. Known genetic variants with effects on important clinical phenotypes, including clopidogrel efficacy and warfarin maintenance dose, highlight the potential translational utility of pharmacogenetic analysis. The emergence of next-generation sequencing offers a promising new tool to explore the links between drug response and genetic variation, both common and rare. To characterize the spectrum of variation in human populations and to evaluate how these differences are linked to drug responses, the National Institutes of Healths Pharmacogenomics Research Network (PGRN) has developed a new platform, PGRNseq. PGRNseq is a low-cost, high-throughput next-generation sequencing platform centered around the custom capture of 84 genes with strong drug phenotype associations. Sequence captured from these genes includes coding regions and 2kb upstream to assess variation within potential regulatory regions. PGRNseqs design includes known variants present on other commercially available pharmacogenetic platforms for backwards compatibility with existing datasets. To test the performance and accuracy of this new tool, we sequenced 32 diverse trios from HapMap and 1000Genomes using the PGRNseq platform. Analysis of Mendelian inconsistencies across test trios identified paralogous regions in which better read mapping and variant calling are needed. In uniquely mapping regions, we found 99.9% genotype concordance at all overlapping sites with orthogonal datasets from HapMap and 1000Genomes. PGRNseq is able to assess known variants of clinical utility such as CYP2C9*3; we found 99.9% genotype concordance at all such sites. Copy number variation is also related to drug response in several genes (e.g. CYP2D6); we are developing methods to discover and type such events using data from PGRNseq. Low-cost, high-throughput platforms such as PGRNseq are needed in order to move genomics forward from the bench to the bedside. Our initial data suggests that PGRNseq could be successfully deployed in a clinical setting to inform patient care and generate large-scale, high-quality data for future genotype-phenotype association studies.
You may contact the first author (during and after the meeting) at