BHIT

BHIT detects high-order epistatic interactions among single nucleotide polymorphisms (SNPs) using a Bayesian partition model with Markov Chain Monte Carlo (MCMC) search to elucidate the genetic architecture of complex traits.

Key Features:

• Bayesian computational framework: Implements a Bayesian partition model capable of analyzing both continuous and discrete data types for case-control studies and quantitative phenotypes.
• High-order interaction detection: Detects complex interactions involving more than two SNPs to capture epistatic effects beyond additive single-marker associations in GWAS.
• MCMC search algorithm: Uses Markov Chain Monte Carlo search techniques to explore the large parameter space of possible interaction configurations.
• Cross-species pipeline: Provides a computational pipeline to apply the method across different species and research use cases.

Scientific Applications:

• Simulation and empirical validation: Applied to simulation data and empirical studies to assess detection of high-order SNP interactions.
• Soybean seed composition: Used to detect high-order interactions associated with soybean oil content and protein content.
• GWAS of complex traits: Applied to genome-wide association study data to investigate epistasis underlying complex traits and diseases, and to inform crop improvement research.
• Comparative performance: Demonstrated improved detection of high-order SNP interactions in comparative analyses versus other available tools.

Methodology:

Employs a Bayesian partition computational method combined with Markov Chain Monte Carlo (MCMC) search; supports analysis of continuous and discrete data for case-control and quantitative phenotypes.