SIGNIFICANT IMPROVEMENT OF SEQUENCE ALIGNMENT CAN BE DONE BY CONSIDERING TRANSITION PROBABILITY BETWEEN TWO CONSECUTIVE PAIRS OF RESIDUES

摘要

It is important to look for accurate sequence alignment because biological sequence analysis depends on the quality of alignment. However it is still a difficult computational problem because of not only the computational complexity but also unsatisfactory objective functions for measuring alignment. Here, we show the significant improvement of sequence alignment can be done by introducing a new metric defined on compound systems of two sequences. In the benchmark tests by PREFAB4 and HOMSTRAD, our method gives higher accuracy than other methods such as Needleman-Wunsch, ClustalW2. Especially for the sequences with less than 30% sequence identity, our method improves the alignment accuracy significantly.