QuickHap: a Quick heuristic algorithm for the single individual Haplotype reconstruction problem

Abstract

Single individual haplotype reconstruction refers to the computational problem of inferring the two distinct copies of each chromosome. Determination of haplotypes offers many advantages for genomic-based studies in various fields of human genetics. Although many methods have been proposed to obtain haplotypes with high accuracy, the rapid and accurate solution of haplotype assembly is still a challenging problem. The largeness of the high-throughput sequence data and the length of the human genome emphasize the importance of the speed of algorithms. In this paper, we propose QuickHap, a heuristic algorithm to achieve a high speed of haplotyping with acceptable accuracy. Our algorithm contains two phases; first, a partial haplotype is built and expanded during several iterations. In this phase, we utilize a new metric to measure the quality of the reconstructed haplotype in each iteration to achieve the optimum solution. The second phase is applied to refine the reconstructed haplotypes to improve accuracy. The result demonstrates that the proposed method can reconstruct the haplotypes with promising accuracy. It outperforms the comparing methods in speed, particularly in dealing with high coverage sequencing data.