Application of clustering technique in multiple sequence alignment
SPIRE'05 Proceedings of the 12th international conference on String Processing and Information Retrieval
| Hi-index | 0.00 |
We present a new method for the calculation of multiple sequence alignments (MSAs). The input to our problem are n protein sequences. We assume that the sequences are related with each other and that there exists some unknown evolutionary tree that corresponds to the MSA. One advantage of our method is that the scoring can be done with reference to this phylogenetic tree, even though the tree structure itself may remain unknown. Instead of computing an evolutionary tree, we only need to compute a circular tour of the tree which is determined via a Traveling Sales-man Problem (TSP) algorithm. Our algorithm can calculate a near optimal MSA and has a performance guarantee of \mathopt (where opt is the optimal score of the MSA). The algorithm runs in \mathtime, where k is the length of the longest input sequence. From there we improve the alignment further. Experimental results are shown at the end.