A high performance fpga-based implementation of position specific iterated blast

Authors:
Server Kasap;Khaled Benkrid;Ying Liu
Affiliations:
University of Edinburgh, Edinburgh, United Kingdom;University of Edinburgh, Edinburgh, United Kingdom;University of Edinburgh, Edinburgh, United Kingdom
Venue:
Proceedings of the ACM/SIGDA international symposium on Field programmable gate arrays
Year:
2009

Citing 1
Cited 3

Processing in Memory: The Terasys Massively Parallel PIM Array

Computer

Bridging the GPGPU-FPGA efficiency gap

Proceedings of the 19th ACM/SIGDA international symposium on Field programmable gate arrays
Towards systolic hardware acceleration for local complexity analysis of massive genomic data

Proceedings of the great lakes symposium on VLSI
FPGA-based hardware acceleration for local complexity analysis of massive genomic data

Integration, the VLSI Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

We present in this paper the first reported FPGA implementation of the Position Specific Iterated BLAST (PSI-BLAST) algorithm. The latter is a heuristic biological sequence alignment algorithm that is widely used in the bioinformatics and computational biology world in order to detect weak homologs. The architecture of our FPGA implementation is parameterized in terms of sequence lengths, scoring matrix, gap penalties and cut-off and threshold values. It is composed of various blmocks each of which performs one step of the algorithm in parallel. This results in high performance implementations, which easily outperform equivalent software implementations by one order of magnitude or more. Furthermore, the core was captured in an FPGA-platform-independent language, namely the Handel-C language, to which no specific resource inference or placement constraints were applied. This makes our core portable across different FPGA families and architectures.