GAMETES

GAMETES generates complex biallelic single nucleotide polymorphism (SNP) n-locus disease models for simulation studies of multi-locus genetic effects, including pure and strict epistasis.

Key Features:

• Generation of pure and strict epistatic models: Produces random, pure, and strictly epistatic n-locus models of biallelic SNPs in which disease risk arises from multi-locus interactions rather than single-locus effects.
• Specification of genetic constraints: Allows explicit specification of heritability, minor allele frequencies (MAFs), and population prevalence for generated models.
• Dataset simulation: Simulates case-control datasets from generated genetic models to create archives of simulated datasets for algorithm testing.
• Support for low heritability scenarios: Can generate lower-heritability models commonly used in algorithm evaluation to assess performance under challenging conditions.
• Random model architectures: Produces random model architectures that represent scenarios where associations are detectable only when all contributing loci are modeled.

Scientific Applications:

• Algorithm evaluation: Benchmarking and evaluating epistasis-detection algorithms such as MDR (Multifactor Dimensionality Reduction) using simulated datasets.
• Method development and testing: Generating test datasets for development and robustness assessment of new computational methods for multi-locus analysis.
• Theoretical studies of genetic interactions: Exploring and characterizing the properties of multi-locus genetic models and epistasis under controlled genetic constraints.

Methodology:

Generates random, purely and strictly epistatic n-locus biallelic SNP models with user-specified heritability, minor allele frequencies, and population prevalence, and simulates datasets from those models to produce archives for algorithm testing.