Search

The evolutionary processes vary among sites of an alignment that strongly affect the accuracy of phylogenetic tree reconstruction. Partitioning an alignment into sub-alignments of sites such that the evolutionary processes at sites in the same sub-alignment are highly similar is a proper strategy. Gene features might be used as reasonable indicators to partition an alignment. However, the gene feature information is not always available or efficient Computational partitioning methods like iterative k-means has been proposed to automatically partition sites into groups based on the similarity of evolutionary rates of sites. Despite obtaining compelling results in terms of AICc and BIC measurements, the k-means method forms a group of all and only invariant sites that results in bias/wrong phylogenetic trees. In this paper, we improve the k-means algorithm by re-classifying invariant sites into different sub-alignments based on their likelihood values. Experimental results on simulated and empirical DNA datasets showed that the new method, called iK-means, overcame the pitfall of the K-means method, consequently, helps improve the quality of the partitioning sub-alignments. We recommend using the iK-means method to level up the accuracy in inferring phylogenetic trees.

Phylogenetic trees inferred from protein sequences are strongly affected by amino acid evolutionary models. Choosing proper models are needed to account for the heterogeneity in evolutionary patterns across sites, especially when analyzing multiple genes or whole genome datasets. Partitioning is a prominent approach to combine sites undergone similar evolutionary processes into separated groups with proper models. The partitioning scheme can be defined by using structural features of the sequences, however, determining structural features of protein sequences is not always practical. Recently, methods have been proposed to automatically cluster sites into groups based on the rates of sites. The rate of sites is a good indicator; however, it is unable to properly reflex the complex evolutionary processes of sites along the protein sequence. In this paper, we present a new algorithm to automatically determine a partitioning scheme based on the best-fit model of sites, i.e., sites belong to the same model will be classified into the same group. Comparing our proposed method with current methods on a set of empirical protein datasets showed that our method helped to build better trees than other methods tested. Our method will significantly improve protein phylogenetic inference from multiple gene or whole genome datasets.

Phylogenetic trees inferred from protein sequences are strongly affected by amino acid substitution models. Although different amino acid substitution models have been proposed, only a few were estimated for specific species such as the FLU model for influenza viruses. Among the most dangerous viruses for human health, dengue is always on top and the cause of dengue fever up to 100 million people per year. In this study, we built a specific amino acid substitution model for dengue protein sequences, called DEN. The dengue protein sequences were obtained from the NCBI dengue database and the model was estimated using the maximum likelihood method. Experiments showed that the new model DEN helped to build better phylogenetic trees than other existing models. We strongly recommend researchers to use the DEN model for analyzing dengue protein data.