Architectures, Software Infrastructures, and Algorithms for Accelerating Large-Scale Genomic Workflows
The overarching goal of this exploratory project is to accelerate biomedical innovation and enable personalized medicine in a cost-effective manner. We will achieve that by (1) identifying sweet-spot genomics architectures that are based on existing electronic components and are accessible via conventional programming models, which achieve performance/density/cost that is >10× superior to that of commodity architecture, and which can be produced in high volume116; (2) accelerating adoption of these architectures by removing the barriers formed by current software infrastructure117; and (3) developing key algorithms118 and their mappings to the sweet-spot architecture for workflows to be accelerated, including disease-associated variant calling, metagenomics, and de novo assembly.