H 3 M 2 : detection of runs of homozygosity from whole-exome sequencing data

• DOI: 10.1093/bioinformatics/btu401
• Publication Date: 2014-06-26T04:26:56Z
• Authors: Alberto Magi; Lorenzo Tattini; Flavia Palombo; Matteo Benelli; Alessandro Gialluisi; Betti Giusti; Rosanna Abbate; Marco Seri; Gian Franco Gensini; Giovanni Romeo; Tommaso Pippucci
• Keywords: Runs of Homozygosity; 1000 Genomes Project; SNP array; dbSNP; SNP

Abstract Motivation: Runs of homozygosity (ROH) are sizable chromosomal stretches homozygous genotypes, ranging in length from tens kilobases to megabases. ROHs can be relevant for population and medical genetics, playing a role predisposition both rare common disorders. commonly detected by single nucleotide polymorphism (SNP) microarrays, but attempts have been made use whole-exome sequencing (WES) data. Currently available methods developed the analysis uniformly spaced SNP-array maps do not fit easily sparse non-uniform distribution WES target design. Results: To meet need an approach specifically tailored data, we H3M2 , original algorithm based on heterogeneous hidden Markov model that incorporates inter-marker distances detect ROH We evaluated performance correctly identify synthetic chromosomes examined its accuracy detecting different (short, medium long) real 1000 genomes project turned out more accurate than GERMLINE PLINK, two state-of-the-art algorithms, especially detection short ROHs. Availability implementation : is collection bash, R Fortran scripts codes freely at https://sourceforge.net/projects/h3m2/ . Contact [e-mail] Supplementary information data Bioinformatics online.