famFLM

famFLM performs regional association analysis in familial genetic data using functional linear models to test associations between quantitative traits and multiple genetic variants within genomic regions.

Key Features:

• Functional Linear Model with Random Polygene Effect: Incorporates a random polygene effect into a functional linear model to test associations between quantitative traits and multiple variants within a genomic region, treating the regional genome as a continuous stochastic function that captures linkage and linkage disequilibrium.
• Familial Data Adaptation: Extends region-based association testing to related individuals by accounting for familial relationships.
• Basis Function Flexibility: Implements B-spline and Fourier basis functions for modeling regional genomic variation.
• Fourier Basis Performance: Comparative analyses reported that Fourier basis models provide superior computational speed and statistical power relative to B-spline or combined bases.
• Statistical Power Enhancement: Shows increased statistical power over burden-based and kernel-based methods across simulation scenarios, including regions with a small number of variants.
• Implementation: Provided as an R-function named famFLM.

Scientific Applications:

• Family-based Regional Association Testing: Identify associations between quantitative traits and genomic regions in family-based genetic studies while accounting for relatedness.
• Rare Variant Detection and Gene Mapping: Detect effects of rare variants and support gene mapping efforts in familial datasets.

Methodology:

Applies functional linear modeling with a random polygene effect, represents genomic regions as continuous stochastic functions expanded using B-spline or Fourier bases, and evaluates performance against burden-based and kernel-based methods via simulations.