Epi-GTBN

Epi-GTBN identifies epistatic loci that influence phenotypic traits by combining genetic tabu algorithms with Bayesian networks to improve detection of gene-gene interactions.

Key Features:

• Bayesian Network Integration: Uses Bayesian networks (BN) as graphical models to represent relationships between genetic loci and phenotypes.
• Genetic Algorithm Enhancement: Incorporates genetic algorithms for rapid global search to explore large solution spaces and avoid local optima.
• Genetic Tabu Algorithm: Integrates a tabu search strategy into crossover and mutation operations to maintain population diversity and accelerate convergence toward global optimal network structures.
• Heuristic Search Strategy: Evolves individual network structures through selection, crossover, and mutation to facilitate discovery of epistatic loci.

Scientific Applications:

• Complex trait epistasis discovery: Detects gene-gene interactions underlying phenotypic traits in complex trait analysis.
• Experimental validation: Demonstrated improved epistasis detection accuracy in comparative studies using simulated and real datasets relative to recent algorithms.

Methodology:

Employs Bayesian networks; uses a genetic algorithm with selection, crossover, and mutation; and integrates tabu search into crossover and mutation to maintain population diversity and accelerate convergence.