Partitioned optimization algorithms for multiple sequence alignment

Multiple sequence alignment is an important and difficult problem in molecular biology and bioinformatics. In this paper, we propose a partitioning approach that significantly improves the solution time and quality by utilizing the locality structure of the problem. The algorithm solves the multiple sequence alignment in three stages. First, an automated and suboptimal partitioning strategy is used to divide the set of sequences into several subsections. Then a multiple sequence alignment algorithm based on ant colony optimization is used to align the sequences of each subsection. Finally, the alignment of original sequences can be obtained by assembling the result of each subsection. The ant colony algorithm is highly optimized in order to avoid local optimal traps and converge to global optima efficiently. Experimental results show that the algorithm can significantly reduce the running time and improve the solution quality on large-scale multiple sequence alignment benchmarks.