BSW: FPGA-accelerated BLAST-Wrapped Smith-Waterman aligner

The current generation of genome sequencers produces orders of magnitude more sequencing data at a fraction of their former cost, a development that has repositioned the sequencing bottleneck from data acquisition to alignment and analysis. Optimal alignment algorithms, such as Smith-Waterman (SW), provide the most desirable output in terms of sensitivity and accuracy, but are perceived as too computationally demanding for practical use. Practical alignment tools, such as NCBI BLAST, have compensated by using increasingly specialized heuristics to provide faster results at the expense of reduced sensitivity and accuracy. This paper presents an FPGA-based alignment tool that considerably accelerates the SW algorithm with a highly configurable and novel hardware computation engine, while providing a robust software interface that seamlessly integrates into existing BLAST pipelines. This new alignment tool, BLAST-wrapped Smith-Waterman (BSW), improves upon previously published SW-acceleration research by providing a refined tool with a familiar and standard BLAST interface and essential local alignment statistics found in production tools. Experimental evaluation of BSW´s performance for several configurations using large, real-world genome datasets shows speedups up to 59 and 7 vs. BLAST (word sizes 7 and 11) and 63 vs. FASTA´s SSEARCH on a single Altera Stratix IV E530 FPGA. Scaled speedups exceeding 3300 vs. SSEARCH were observed when executed on up to 64 FPGAs of Novo-G, the reconfigurable supercomputer in the NSF CHREC Center at Florida.