KM Prakash Lingam, S Chandrakala
Anna University of Technology, Dept. of Computer Science, Velammal Engineering College, Chennai, Tamil Nadu, India.
A Multiple sequence alignment is a critical step towards comparing sequence similarities and revealing genetic relationships among evolutionarily related species. The Multiple Sequence Alignment (MSA) problem is the basis for many bioinformatics applications. It can be used to infer evolutionary history or to discover conserved regions among closely related protein or DNA sequences. It can provide information about the evolutionary history of the respective sequences. It can give insight into the basis for sequence similarities between homologous sequences. Homologous sequences are sequences that share some evolutionary origin. Progressive alignments based on Dynamic Programming (DP) are by far the most widely used heuristic multiple sequence alignment method. This approach has the great advantages of speed and simplicity combined with reasonable sensitivity, even though it does not guarantee any level of optimization. The progressive alignment consists of three steps. First, it computes the similarities of all possible pair-wise sequence alignment using sequence alignment algorithm. Then, it constructs a guide tree according to the similarities. Finally, it merges the sequences into a multiple alignment. Iterative method is another technique for MSA. It refines the alignment through a series of iterations until no more improvements can be made. These algorithms use a iterative approach in which the existing alignments can be realigned during the addition of more sequences to the multiple sequence alignment. They involve extracting sequences one by one from a multiple alignment and realigning them to the remaining sequences. The procedure is terminated when no more improvement can be made. Iterative methods take more time to construct a multiple sequence alignment. This paper attemps to give an overview of recent approaches in multiple sequence Alignment methods and the challenges ahead like aligning large data sets, sequences with high dissimilarity etc. Read more…