Haplotype Inference on Pedigrees with Recombinations and Mutations

Abstract

Haplotype Inference (HI) is a computational challenge of crucial importance in a range of genetic studies, such as functional genomics, pharmacogenetics and population genetics. Pedigrees have been shown a valuable data that allows us to infer haplotypes from genotypes more accurately than population data, since Mendelian inheritance restricts the set of possible solutions. In order to overcome the limitations of classic statistical haplotyping methods, a combinatorial formulation of the HI problem on pedigrees has been proposed in the literature, called Minimum-Recombinant Haplotype Configuration (MRHC) problem, that allows a single type of genetic variation events, namely recombinations. In this work, we define a new problem, called Minimum-Change Haplotype Configuration (MRHC), that extends the MRHC formulation by allowing also a second type of natural variation events: mutations. We propose an efficient and accurate heuristic algorithm for MRHC based on an L-reduction to a well-known coding problem. Our heuristic can also be used to solve the original MRHC problem and it can take advantage of additional knowledge about the input genotypes, such as the presence of recombination hotspots and different rates of recombinations and mutations. Finally, we present an extensive experimental evaluation and comparison of our heuristic algorithm with several other state-of-the-art methods for HI on pedigrees under several simulated scenarios.