logicFS

logicFS identifies and quantifies interactions among binary variables, particularly single nucleotide polymorphisms (SNPs), using logic regression to detect combinatorial effects associated with complex diseases such as sporadic breast cancer.

Key Features:

• Logic regression for interaction identification: logicFS employs logic regression to model and identify combinatorial interactions between SNPs.
• Quantification of SNP interactions: logicFS implements two novel measures to directly quantify the contribution of SNP combinations to disease status and assess interaction significance.
• Bagging for enhanced classification: logicFS provides a bagging (bootstrap aggregating) version of logic regression to reduce variance and improve classification stability.

Scientific Applications:

• Genetic epidemiology of complex diseases: identifying SNP interactions that contribute to multifactorial conditions such as sporadic breast cancer.
• Prioritization for risk assessment: quantifying interaction importance to prioritize SNP combinations for risk assessment and follow-up studies.

Methodology:

Applies logic regression to genetic case-control data, includes a bagged logic regression variant, was demonstrated on simulated datasets and on real-world SNP data from the GENICA study, and uses two measures to quantify interaction importance.