3-Way Alignment can Improve Multiple Sequence Alignment of Highly Diverged Sequences

Abstract

The standard approach for constructing a phylogenetic tree from a set of sequences consists of two key stages. First, Multiple Sequence Alignment (MSA) of the sequences is computed. The aligned data is then used to reconstruct the phylogenetic tree. The accuracy of the final tree heavily relies on the quality of MSA. Consequently, obtaining an accurate MSA is critical for improving the accuracy of the phylogenetic information. The accuracy of MSA decreases with sequence divergence. Numerous studies have demonstrated the difficulty in obtaining good MSA for reconstructing deep phylogeny, e.g. arthropod phylogeny. The accuracy of MSA is universally recognized as the limiting factor for tracing evolutionary history back in time. In this thesis, we assess the effect of guide tree on the accuracy of Multiple Sequence Alignment and subsequent phylogenetic tree and apply 3-way alignment to generate more accurate guide trees. We evaluate the performance of MAFFT on simulated dataset with different level of divergency including highly diverged sequences. Our results show that 3-way alignment outperforms the MAFFT's default method in generating more accurate guide trees specially for highly diverged sequences. Moreover, MSAs generated by using 3-way alignment guide trees lead to more accurate subsequent phylogenetic trees compared to MSAs generated by MAFFT default method. In addition, we showed the significant role of a good guide tree in improving final phylogenetic trees. Finally, our results emphasis on the significantly low performance of MAFFT in aligning highly diverged sequences.