WAFFECT

WAFFECT simulates phenotypic datasets under specified disease models for genome-wide association (GWA) studies as an R package, maintaining a constant total number of cases across simulations to assess statistical power to detect susceptibility variants.

Key Features:

• Simulation Algorithms: Implements three algorithms—a rejection algorithm, a numerical Markov chain Monte Carlo (MCMC) approach, and an exact backward sampling algorithm—for simulating phenotypes without generating new genotypes.
• Efficiency and Speed: The exact backward sampling algorithm provides a dramatic speed advantage over the other methods, facilitating large-scale simulations.
• Validation and Consistency: Algorithms have been validated on simulated and realistic datasets and show consistency with established methods such as Hapgen.
• Flexibility in Disease Modeling: Supports an arbitrary number of susceptibility SNPs, inclusion of epistatic effects, gene–environment interactions, and hybrid genetic models without requiring haplotype frequencies or recombination rates.
• Application Example: Applied to 1000 Genomes Project data comprising 629 individuals (314 cases) and 8,048 SNPs on chromosome X using an additive disease model with two susceptibility SNPs and an epistatic effect.

Scientific Applications:

• Power Analysis in GWA Studies: Empirically estimates the power to detect genetic variants associated with disease in GWA studies.
• Disease Model Exploration: Enables exploration of diverse genetic models, including epistasis and gene–environment interactions, without genotype-modeling constraints.
• Impact of Disease Prevalence: Assesses how disease prevalence affects the performance and power of GWA studies.

Methodology:

Simulates phenotypes directly under a specified disease model while maintaining a fixed total number of cases, using rejection sampling, numerical MCMC, and an exact backward sampling algorithm, and does not generate new genotypes.