Smith-Waterman Acceleration on Intel's FPGA with OpenCL for Long DNA Sequences

SWIFOLD is a software to accelerate Smith-Waterman alignment of long DNA sequences on Intel's FPGAs.

We have executed SWIFOLD in different FPGA-based platforms. Some of the performance results are detailed below.

Experimental platform: A server with two Intel Xeon CPU E5-2670 8-core 2.60GHz, 48 GB main memory and an Intel Arria 10 GX with 2 GByte DDR3. Intel’s ICC compiler (version 17.0.1.132). Intel FPGA OpenCL SDK 16.0

Experimental platform: A server with two Intel Xeon CPU E5-2670 8-core 2.60GHz, 32 GB main memory and an Altera Stratix V GSD5 Half-Length PCIe Board with Dual DDR3 (two banks of 4 GByte DDR3). Intel’s ICC compiler (version 16.0.3). Quartus II DKE V12.0 2 with OpenCL SDK v14.0.

./swifold -i seq1.fasta -j -seq2.fasta

  -i,   --input=<string> Query sequence filename (must be in FASTA format) [REQUIRED]
  
  -j,   --target=<string> Target sequence filename (must be in FASTA format) [REQUIRED]
  
  -M,   --match=<integer> Match score (default: 1)

  -X,   --mismatch=<integer> Mismatch score (default: 3)
  
  -g,   --gap_open=<integer> Gap open penalty (default: 5)

  -e,   --gap_extend=<integer> Gap extend penalty (default: 2)

  -?,   --help Give this help list
        --usage Give a short usage message

Remember that both performance and resource usage depend on the BLOCK_WIDTH constant. Tune BLOCK_WIDTH according to your needs (see swifold/device/sw.cl).

Rucci E., Garcia C., Botella G., De Giusti A., Naiouf M., Prieto-Matias M. (2018) SWIFOLD: Smith-Waterman Acceleration on Intel’s FPGA with OpenCL for Long DNA Sequences. BMC Systems Biology2018 12(Suppl 5):96. DOI https://doi.org/10.1186/s12918-018-0614-6

Rucci E., Garcia C., Botella G., De Giusti A., Naiouf M., Prieto-Matias M. (2017) Accelerating Smith-Waterman Alignment of Long DNA Sequences with OpenCL on FPGA. In: Rojas I., Ortuño F. (eds) Bioinformatics and Biomedical Engineering. IWBBIO 2017. Lecture Notes in Computer Science, vol 10209. Springer, Cham. DOI https://doi.org/10.1007/978-3-319-56154-7_45

• January 25, 2019 [v1.0] Source code released
• September 01, 2017 Binary code released

If you have any question or suggestion, please contact Enzo Rucci (erucci [at] lidi.info.unlp.edu.ar)