HAPCOL: HAPlotype Assembly COrrecting Long Reads

A fast and memory-efficient method for haplotype assembly from long gapless reads, like those produced by SMRT sequencing technologies (PacBio RS II) and Oxford Nanopore flow cell technologies (MinION). HapCol implements a fixed-parameter algorithm for the k-constrained Minimum Error Correction problem (k-cMEC), a variant of the well-known MEC problem where the maximum number of corrections per column is bounded by an integer k. HapCol, while is as accurate as other exact state-of-the-art combinatorial approaches, is significantly faster and more memory-efficient than them. Moreover, HapCol is able to process datasets composed of both long reads (over 100 000bp long) and coverages up to 25x on standard workstations/small servers, whereas the other approaches cannot handle long reads or coverages greater than 20x. The detailed description of the algorithm, along with an experimental comparison with other state-of-the-art haplotype assembly tools, is presented in: Yuri Pirola, Simone Zaccaria, Riccardo Dondi, Gunnar W. Klau, Nadia Pisanti, and Paola Bonizzoni HapCol: Accurate and Memory-efficient Haplotype Assembly from Long Reads. Bioinformatics. doi:10.1093/bioinformatics/btv495