NGS Based Haplotype Assembly Using Matrix Completion

TL;DR

This paper introduces novel matrix completion algorithms for haplotype assembly from NGS reads, demonstrating improved performance over existing methods in terms of SNP missing rate and haplotype block length.

Contribution

The paper develops three new algorithms, HapSVT, HapNuc, and HapOPT, applying matrix completion techniques to enhance haplotype assembly accuracy and efficiency.

Findings

HapOPT outperforms HapCUT2 in SNP missing rate and haplotype block length.

The algorithms achieve comparable reconstruction and switch error rates to state-of-the-art methods.

The MATLAB implementation is publicly available for research use.

Abstract

We apply matrix completion methods for haplotype assembly from NGS reads to develop the new HapSVT, HapNuc, and HapOPT algorithms. This is performed by applying a mathematical model to convert the reads to an incomplete matrix and estimating unknown components. This process is followed by quantizing and decoding the completed matrix in order to estimate haplotypes. These algorithms are compared to the state-of-the-art algorithms using simulated data as well as the real fosmid data. It is shown that the SNP missing rate and the haplotype block length of the proposed HapOPT are better than those of HapCUT2 with comparable accuracy in terms of reconstruction rate and switch error rate. A program implementing the proposed algorithms in MATLAB is freely available at https://github.com/smajidian/HapMC.