Enhancement of proclust algorithm using parallel methods for protein sequence clustering
Loading...
Date
2008
Authors
Haj Assayony, Mohammed Omer
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Graph-based protein sequences clustering algorithms produce good results in clustering
protein sequences into accurate cluster. The main advantage of these algorithms is that
they convert the clustering problem into a well-known problem in graph theory and use
the algorithms and techniques used in graph theory to deal with it. The drawback of
these algorithms is that they exploit long time in producing the results. This research is
about studying these algorithms and designing parallel methods and techniques to
improve the performance of these algorithms. The research concentrates on designing
parallel methods for improving the performance of one of these algorithms: ProClust
which is depending on the concept of finding strongly connected components of a
directed graph. The methods and techniques that we have designed for improving
ProClust are suggested to be vsed in other graph-based protein sequence clustering. The
suggested parallel methods distribute the computational work of the algorithm almost
evenly among a set of interconnected processors such that the communications among
them are low comparing with the computational load. The results of implementing the
parallel method used in improving the main step in ProClust, which is the process of
finding strongly connected components of the directed graph, are acceptable comparing
with the results obtained from other works.
Description
Keywords
Proclust algorithm , Sequence clustering