Comparison of multiple DNA alignment algorithms for Labiatae molecular phylogeny inferences

Multiple alignment of sets of nucleotide or amino acid sequences are usually required prior to phylogenetic studies for phylogeny inferences. There are several widely used programs for carrying out automatic multiple alignment of nucleotide sequences [1,2,3]. Since there exist different type of multiple alignment programs, the selection of proper alignment program is important for true phylogeny inferences. In order to compare different alignment algorithms implemented in five different multiple alignment programs (ClustalW, T-Coffee, MAFFT, Kalign and MUSCLE, all implemented in the Ugene software) were used to align internal transcribed spacer sequences of 24 taxa of Labiatae family. MetaPIGA 2.0 software was used to obtain phylogenetic trees. Analysis parameters of MetaPIGA 2.0 were set to heuristic search using stochastic consensus pruning (metaGA), GTR model of likelihood rate test and other parameters were kept at default values of the program. Evaluations of the multiple alignment algorithms were based on bootstrap values and groupings at the genus level of consensus trees. Analyses indicated that phylogeny inferences were affected with the multiple alignment algorithms used. Among the tested multiple alignment programs, sequences aligned with MAFFT produced better phylogenetic tree. Further studies will be useful to reveal whether the use of different phylogeny programs and different gene sequences could overcome the effects of multiple alignments on phylogeny inferences.

Acknowledgement: This research is supported by the Scientific Research Projects Coordination Unit of Akdeniz University.

References: 1. Ince AG et al. (2005) Akdeniz Univ Ziraat Fak Derg 18: 157–162. 2. Helaers R, Milinkovitch MC (2010) BMC Bioinformatics 11: 379. 3. Ince AG et al. (2010) Genet Resour Crop Ev 57: 773–779.