CAREER: System Support for Scalable, Fast, and Power-Efficient Genome Sequencing

点击翻译按钮获取中文摘要

Genome sequencing is the key to enhancing public health, optimizing personality medicine, and facilitating viral surveillance. However, the current genome sequencing pipeline lacks full-stack system support and consumes tens or even hundreds of CPU hours, causing delays in treatment, patient deterioration, or even mortality. This project integrates four synergistic research tasks to enable scalable, fast, and power-efficient genome sequencing: (1) This project will create a benchmark suite to identify the performance/power bottlenecks in the pipeline; (2) This project presents a scalable algorithm for the most time-consuming Seed-and-Extension operations in the genome sequencing pipeline; (3) This project will design an ultra-low power processing-in-memory architecture for accelerating intensive operators in the genome sequencing pipeline; (4) This project will develop an automated compilation framework to deploy the genome sequencing pipeline to a CPU/accelerator heterogeneous system. The outcomes of this proposal will lower the barrier to entry for genome sequencing development, equipping the next generation of users, developers, entrepreneurs, and scientists to use genome sequencing, without help from software and hardware experts. The research will inspire future computer systems and computational biology research for specialized genome sequencing pipeline acceleration. This proposal integrates research and education including graduate and undergraduate curriculum development and research mentoring, a high school intern program to enable early learners to use genome sequencing and develop an interest in STEM.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.