Performance evaluation of leading protein multiple sequence alignment methods

Abstract

Protein Multiple sequence alignment (MSA) is a process, that helps in alignment of more than two protein sequences to establish an evolutionary relationship between the sequences. As part of Protein MSA, the biological sequences are aligned in a way to identify maximum similarities. Over time the sequencing technologies are becoming more sophisticated and hence the volume of biological data generated is increasing at an enormous rate. This increase in volume of data poses a challenge to the existing methods used to perform effective MSA as with the increase in data volume the computational complexities also increases and the speed to process decreases. The accuracy of MSA is another factor critically important as many bioinformatics inferences are dependent on the output of MSA. This paper elaborates on the existing state of the art methods of protein MSA and performs a comparison of four leading methods namely MAFFT, Clustal Omega, MUSCLE and ProbCons based on the speed and accuracy of these methods. BAliBASE version 3.0 (BAliBASE is a repository of manually refined multiple sequence alignments) has been used as a benchmark database and accuracy of alignment methods is computed through the two widely used criteria named Sum of pair score (SPscore) and total column score (TCscore). We also recorded the execution time for each method in order to compute the execution speed.