DHSMAP

DHSMAP estimates linkage disequilibrium and fine-maps genetic variants for qualitative traits using a multilocus Decay of Haplotype Sharing (DHS) model.

Key Features:

• Multilocus Model: Employs a multilocus model to analyze densely mapped haplotype data by modeling dependence among multiple tightly linked loci.
• DHS Parameter: Quantifies LD as the expected genetic distance over which the ancestral haplotype is preserved and can be interpreted as the inverse of the time in generations to the ancestral haplotype.
• Handling Complex Population Structures: Accommodates multiple origins of alleles and mutations and addresses missing observations and haplotype phase ambiguities using a hidden Markov model.
• Estimation Methods: Estimates LD by maximum likelihood for tractable population structures and, for more complicated scenarios, computes model-derived means and covariances and solves a quasi-score estimating equation.
• Simulation Validation: Performance and fine-mapping accuracy have been evaluated using extensive simulations.

Scientific Applications:

• Gene Mapping by LD: Applied to fine-map genetic variants associated with qualitative traits through linkage disequilibrium analysis.
• Medical Genetics Case Studies: Demonstrated on published datasets including cystic fibrosis and progressive myoclonus epilepsy.

Methodology:

Multilocus Decay of Haplotype Sharing (DHS) model; DHS parameter as expected preserved genetic distance/inverse generations; hidden Markov model to handle missing data and haplotype ambiguity; estimation by maximum likelihood or by solving a quasi-score estimating equation using model-derived means and covariances; simulation-based validation.