A study of pClust settings

Abstract

Recently, high-throughput approaches to DNA sequencing such as massive parallel sequencing have resulted in the availability of a vast number of whole genome sequences. This availability has presented scientists with an unprecedented opportunity to gain knowledge by means of datamining and data analysis. A number of our datamining and data analysis strategies are based on the use of a fast and accurate software tool, pClust , to group protein sequences into homologous clusters. However, pClust has a number of parameters with values that must be chosen, and the choice of these values affects the accuracy of the clustering results. In this paper, we present a study of the most significant parameters: alignment length, match similarity, and optimal score. In addition, we study both local and semi-global alignments.