SWPS3 – fast multi-threaded vectorized Smith-Waterman for IBM Cell/B.E. and ×86/SSE2
2008

SWPS3: A Fast Implementation of the Smith-Waterman Algorithm

Sample size: 359942 publication Evidence: high

Author Information

Author(s): Adam Szalkowski, Christian Ledergerber, Philipp Krähenbühl, Christophe Dessimoz

Primary Institution: ETH Zürich

Hypothesis

Can the Smith-Waterman local alignment algorithm be optimized for better performance on different architectures?

Conclusion

The SWPS3 implementation demonstrates that the Cell/BE can effectively align biological sequences, achieving performance levels comparable to heuristic methods.

Supporting Evidence

  • SWPS3 is the fastest implementation of the Smith-Waterman algorithm on the Cell/BE, outperforming previous implementations by a factor of at least 4.
  • On a quad-core Intel Pentium, SWPS3 achieves up to 15.7 billion cell-updates per second.
  • SWPS3's performance is comparable to heuristic methods like BLAST for long protein sequences.

Takeaway

This study shows a new tool that helps compare biological sequences really fast, making it easier for scientists to understand how they are related.

Methodology

The study involved benchmarking the SWPS3 implementation against other alignment tools using protein sequences and various scoring matrices.

Limitations

SWPS3 can only compute local protein sequence alignments with affine gaps and does not display the resulting alignment.

Digital Object Identifier (DOI)

10.1186/1756-0500-1-107

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication