ROHMM

ROHMM applies a flexible Hidden Markov Model to detect runs of homozygosity (ROH) from genotype probability data, including analysis of non-pseudoautosomal regions of the X chromosome in next-generation sequencing datasets.

Key Features:

• Methodology: Utilizes genotype probabilities within an HMM framework to identify ROHs and can be applied to non-pseudoautosomal regions of the X chromosome for NGS data.
• Implementation: Implemented in Java.
• Performance: Evaluated on simulated datasets, the 1000 Genomes Project population data, and clinical samples, reporting robust and accurate homozygosity estimation and improved ROH detection relative to existing methods.

Scientific Applications:

• Genetic Epidemiology: Identifies populations with elevated consanguinity to inform studies of recessive genetic disorder prevalence.
• Clinical Genetics: Pinpoints regions of homozygosity that may harbor recessive deleterious mutations for disease diagnosis and interpretation.
• Population Genetics: Detects ROHs to analyze genetic diversity, population structure, evolutionary history, and demographic events.

Methodology:

ROHMM uses genotype probabilities input into a Hidden Markov Model to call ROHs and supports analysis of non-pseudoautosomal X chromosome regions in next-generation sequencing data.