Massively Parallel Biosequence Analysis

Richard Hughey
Computer Engineering Board of Studies
University of California, Santa Cruz
rph@ce.ucsc.edu
(408) 459-2939

Technical Report UCSC-CRL-93-14
April 2, 1993

Abstract:

Massive parallelism is required for the analysis of the rapidly growing biosequence databases. First, this paper compares and benchmarks methods for dynamic programming sequence analysis on several parallel platforms. Next, a new hidden Markov model method and its implementation on several parallel machines is discussed. Finally, the results of a series of experiments using this massively parallel implementation are described.

Keywords: Computational biology, dynamic programming, parallel algorithms, systolic co-processors, hidden Markov models.

• Introduction
  • Biosequence analysis methods
  • Hardware for biosequence analysis
• Hidden Markov model sequence analysis
• Massively Parallel Implementation
• Evaluation
• Biological Results
• Acknowledgments
• References
• About this document ...