FPGA-Accelerated 3rd Generation DNA Sequencing

被引:10
|
作者
Wu, Zhongpan [1 ]
Hammad, Karim [1 ,2 ]
Ghafar-Zadeh, Ebrahim [1 ]
Magierowski, Sebastian [1 ]
机构
[1] York Univ, Dept Elect Engn & Comp Sci, Toronto, ON M3J 1P3, Canada
[2] Arab Acad Sci Technol & Maritime Transport, Dept Elect & Commun Engn, Cairo, Egypt
基金
加拿大自然科学与工程研究理事会;
关键词
DNA; Sensors; Sequential analysis; Nanobioscience; Biomedical measurement; Hidden Markov models; Field programmable gate arrays; Basecalling; DNA sequencing; FPGA acceleration; HMM; nanopore; PCIe; RIFFA;
D O I
10.1109/TBCAS.2019.2958049
中图分类号
R318 [生物医学工程];
学科分类号
0831 ;
摘要
DNA measurement machines are undergoing an orders-of-magnitude size and power reduction. As a result, the analysis of genetic molecules is increasingly appropriate for mobile platforms. However, sequencing these measurements (converting to the molecule's A-C-G-T text equivalent) requires intense computing resources, a problem for potential realizations as mobile devices. This paper proposes a step towards addressing this issue, the design and implementation of a low-power real-time FPGA hardware accelerator for the basecalling task of nanopore-based DNA measurements. Key basecalling computations are identified and ported to a custom FPGA which operates in tandem with a CPU across a high-speed serial link and a simple API. A measured speed-up over CPU-only basecalling in excess of 100X is realized with an energy efficiency improvement of three orders of magnitude.
引用
收藏
页码:65 / 74
页数:10
相关论文
共 50 条