FREGAT

FREGAT applies functional linear regression to perform gene-based association analysis of complex traits by integrating genetic variant positions and allele-frequency-derived weights to prioritize causal variants.

Key Features:

• Functional Linear Regression Models: Employs functional linear regression models to analyze quantitative traits by combining individual genetic variant information while accounting for positional relationships.
• Support for Study Designs: Applicable to both family-based and population (independent) samples for gene-based association testing.
• Weighted Components: Introduces allele-specific weights derived from allele frequencies modeled via a beta distribution to prioritize more informative components and causal variants.
• Real-data Association Example: Detected an association between diastolic blood pressure and the VMP1 gene in the ORCADES sample with P = 8.18×10^-6.
• Validation and Performance: Validated through simulations based on GAW17 genotypes, showing controlled type I error rates and that weighted models yield lower p-values for known genes compared to unweighted analyses.

Scientific Applications:

• Gene-based association mapping: Identifying associations between genetic variants and complex traits.
• Quantitative trait analysis: Analyzing quantitative phenotypes in both family-based and population cohort studies.
• Method validation and power assessment: Evaluating type I error and power of gene-based tests via simulations using GAW17 genotypes.

Methodology:

Uses functional linear regression models with allele-specific weights defined from allele frequencies modeled by a beta distribution; performance evaluated by simulations on GAW17 genotypes and by real-data testing in the ORCADES sample (diastolic blood pressure–VMP1 association).