High Performance Computing System for Human Genomics

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): High performance computation lies at the heart of modern human statistical genomics. In the Department of Genetics at the Southwest Foundation for Biomedical Research (SFBR), we specialize in the genetic dissection of human complex diseases that typically involve the actions of many genes and environmental factors. For a given genetic project on complex diseases, we typically generate vast amounts of sequence- based genetic, transcriptomic, and disease-related phenotypic information. Because family-based gene localization and identification studies are substantially more powerful than those undertaken in unrelated individuals, we also focus on studies involving very large extended pedigrees. Statistical genetic theory can be invoked to show that such a design is optimal when rare genetic variants are likely to play a role in disease risk. Accumulating results from genome-wide association studies, which are capable of only finding common disease-related genetic variants, now strongly suggest that rare variants are of substantial importance in human complex disease. However, the utilization of extended pedigree information incurs a strong computational price, since family-based data are non-independent and therefore must be analyzed simultaneously. These issues all lead to the need for ever more powerful computational technologies. In this application, we request funds to purchase a large Opteron-based high performance parallel computing cluster (an M&A 5000 Core Computational Cluster) to specifically aid ongoing research projects with more than $81 million dollars of NIH funding. PUBLIC HEALTH RELEVANCE: The Department of Genetics at the Southwest Foundation for Biomedical Research focuses on understanding the genetic factors involved in human common complex diseases, including cardiovascular disease, diabetes, obesity, aging, pre-eclampsia, psychiatric disorders and infectious diseases. Investigators employ state-of-the science statistical/computational methods to process vast amounts of sequence-based genetic, transcriptomic, and disease-related phenotypic data to localize and identify novel genes influencing disease risk. Acquisition of a high performance parallel computing cluster (the M&A 5000 Core Computational Cluster) will speed the pace of gene discovery in such disorders. The proposed equipment will also significantly enhance both job creation and job retention at the Southwest Foundation for Biomedical Research and M&A Technology, Inc.