Local DNA Sequence Alignment in a Cluster ofWorkstations: Algorithms and Tools

Alba Cristina M. A. MeloMaria Emilia M.T. WalterRenata Cristina F. MeloMarcelo N. P. SantanaRodolfo B. Batista

Distributed Shared Memory systems allow the use of theshared memory programming paradigm in distributed architectureswhere no physically shared memory exist. Scopeconsistent software DSMs provide a relaxed memory modelthat reduces the coherence overhead by ensuring consistencyonly at synchronization operations, on a per-lock basis.Much of the work in DSM systems is validated by benchmarksand there are only a few examples of real parallelapplications running on DSM systems. Sequence comparisonis a basic operation in DNA sequencing projects, andmost of sequence comparison methods used are based onheuristics, that are faster but do not produce optimal alignments.Recently, many organisms had their DNA entirelysequenced, and this reality presents the need for comparinglong DNA sequences, which is a challenging task due toits high demands for computational power and memory. Inthis article, we present and evaluate a parallelization strategyfor implementing a sequence alignment algorithm forlong sequences. This strategy was implemented in JIAJIA,a scope consistent software DSM system. Our results onan eight-machine cluster presented good speedups, showingthat our parallelization strategy and programming supportwere appropriate.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: