hapassoc

hapassoc performs likelihood-based inference of trait associations with SNP haplotypes and environmental cofactors using the Expectation-Maximization (EM) algorithm.

Key Features:

• EM Algorithm Implementation: Employs the Expectation-Maximization (EM) algorithm for maximum-likelihood inference of haplotype–trait associations.
• Handling Missing SNP Genotype Information: Accommodates missing SNP genotype data within the EM framework to enable inference from incomplete datasets.
• Generalized Linear Models (GLMs): Integrates haplotype and environmental cofactor analysis within generalized linear models for association testing.
• Standard Error Calculation: Implements standard error calculation methods that differ from previous implementations by Lake et al.
• Theoretical Basis: Implements methodologies described by Burkett et al. (2004).
• Dependence Robustness: Analysis is robust to low levels of dependence between haplotypes and non-genetic factors but may produce biased inference under moderate to strong dependence.

Scientific Applications:

• Genetic Epidemiology: Perform haplotype-based association analyses in genetic epidemiology studies to relate genetic variation to traits.
• Marker Discovery: Identify genetic markers and haplotypes associated with complex diseases and phenotypic traits.
• Gene–Environment Interaction: Assess interactions between haplotypes and environmental cofactors within a GLM framework.
• Analysis with Incomplete Data: Conduct statistical association analyses when SNP genotype data are partially missing.

Methodology:

Applies the Expectation-Maximization (EM) algorithm for maximum-likelihood estimation within generalized linear models, accommodates missing SNP genotype information, integrates haplotype and environmental cofactor data, and computes standard errors using methods distinct from Lake et al.; robustness depends on the level of dependence between haplotypes and non-genetic factors.