FRGEpistasis

FRGEpistasis performs epistasis analysis of rare genetic variants in next-generation sequencing (NGS) data by applying functional regression models to test interactions between genomic regions.

Key Features:

• Functional Regression Model: Implements a functional regression framework that evaluates interactions between sets of loci or genomic regions rather than individual loci.
• Collective SNP-Pair Testing: Tests interactions by collectively evaluating all possible SNP pairs between two specified genomic regions.
• High-Dimensional Data Reduction and Functional Data Analysis: Employs data reduction techniques and functional data analysis to handle high-dimensional NGS and exome sequencing datasets.
• Statistical Validation: Demonstrated correct type I error control and improved interaction detection relative to traditional pairwise interaction analyses through intensive simulations.
• Application to Exome Sequencing Data and Multiple Testing Correction: Applied to NHLBI's Exome Sequencing Project (ESP) and CHARGE-S exome data with significance assessed using Bonferroni correction.

Scientific Applications:

• Epistasis detection in rare variants: Identifies gene–gene interactions involving rare variants across genomic regions in NGS and exome sequencing studies.
• Exome-sequence studies (ESP and CHARGE-S): Applied to ESP and CHARGE-S exome data to discover and replicate significant gene–gene interaction pairs.
• Genetic architecture of complex traits: Supports investigation of how interactions among rare variants contribute to complex phenotypes and diseases.

Methodology:

Uses a functional regression model with high-dimensional data reduction and functional data analysis to collectively test all SNP pairs between two genomic regions; performance assessed by simulation for type I error and detection power and results evaluated with Bonferroni correction in exome sequencing data.