Dataflow Acceleration of Smith-Waterman with Traceback for High Throughput Next Generation Sequencing

Abstract

Smith-Waterman algorithm is widely adopted by most popular DNA sequence aligners. The inherent algorithm computational intensity and the vast amount of NGS input data it operates on, create a bottleneck in genomic analysis flows for short-read alignment. FPGA architectures have been extensively leveraged to alleviate the problem, each one adopting a different approach. In existing solutions, effective co-design of the NGS short-read alignment still remains an open issue, mainly due to narrow view on real integration aspects, such as system wide communication and accelerator call overheads. In this paper, we propose a dataflow architecture for Smith-Waterman Matrix-fill and Traceback alignment stages, to perform short-read alignment on NGS data. The architectural decision of moving both stages on chip extinguishes the communication overhead, and coupled with radical software restructuring, allows for efficient integration into widely-used Bowtie2 aligner. This approach delivers ×18 speedup over the respective Bowtie2 standalone components, while our codesigned Bowtie2 demonstrates a 35% boost in performance.

An FPGA Accelerator of the Wavefront Algorithm for Genomics Pairwise Alignment

“…For this reason, some accelerators proposed in the literature are limited to compute the alignment score [54], [55], but not the CIGAR. In contrast, other designs offer more flexibility and allow the computation of the full CIGAR alignment at the expense of lower performance [31], [56], [63].…”

“…GCUPS OF DIFFERENT SWG FPGA ACCELERATED METHODS. GCUPs for this accelerator are not reported explicitly, we calculate them from the data provided in the original paper[56]. ** This accelerator does not perform backtrace.…”

On Predictable Reconfigurable System Design

“…Smith-Waterman with Traceback for High Throughput Next Generation Sequencing: The accelerated DNA sequencing aligner developed in Reference [41] used the methodology. The analysis (Part 1) and performance prediction (Part 3) of the methodology identified early that communication overheads could negate any alignment acceleration of the entire application.…”

Genome Sequence Alignment - Design Space Exploration for Optimal Performance and Energy Architectures

“…The GenAx accelerator [34] provides around 31.7x speedup as compared to 14-core Xeon E5 server. The authors in [35] implement a dataflow architecture on FPGA for Smith-Waterman Matrix-fill and Traceback stages, widely used in sequence aligners like BWA-MEM [21] and Bowtie2 [20]. However, this only accelerates the Smith-Waterman part of the sequence alignment application and the other stages of the application have to be run on the CPU.…”