A combination of PSO and k-means methods to solve haplotype reconstruction problem

Disease association study is of great importance among various fields of study in bioinformatics. Computational methods happen to be advantageous specifically when experimental approaches fail to obtain accurate results. Haplotypes are believed to be the most responsible biological data for genetic diseases. In this paper, the problem of reconstructing haplotypes from error-containing SNP fragments is discussed. For this purpose, two new methods have been proposed by a combination of k-means clustering and particle swarm optimization algorithm. The methods and their implementation results on real biological and simulation datasets are represented which shows that they outperform the methods used alone.