ROHan

ROHan estimates heterozygosity and detects runs of homozygosity across genomes to enable accurate assessment of genetic diversity in modern and ancient samples.

Key Features:

• Probabilistic Methodology: Integrates a local Bayesian model with a genome-wide Hidden Markov Model (HMM) to detect heterozygous sites and delineate runs of homozygosity.
• Application to Modern and Ancient Samples: Handles modern and ancient DNA while accounting for low sequencing coverage and post-mortem DNA damage.
• Enhanced Accuracy in Heterozygosity Estimation: Distinguishes true heterozygous sites from sequencing errors and adjusts heterozygosity estimates outside runs of homozygosity.
• Detection of Large Runs of Homozygosity: Identifies extensive runs of homozygosity spanning megabase scales, relevant for highly inbred individuals and domesticated species.
• Reliable Confidence Intervals: Produces confidence intervals for genome-wide heterozygosity that are reliable at low coverage (≈5–6× for modern samples; ≈7–8× for ancient samples with moderate damage).

Scientific Applications:

• Population Genetics: Estimates genetic diversity and inbreeding levels within and between populations.
• Evolutionary Biology: Informs reconstruction of evolutionary histories through heterozygosity-based analyses.
• Domestication and Inbreeding Studies: Characterizes long runs of homozygosity and inbreeding in domesticated animals.

Methodology:

Employs a dual-layered probabilistic framework: a local Bayesian model for per-window heterozygosity estimation that accounts for sequencing errors and coverage variability, and a genome-wide HMM that models transitions between homozygous and heterozygous states to identify runs of homozygosity.