A Constrained Generalized Functional Linear Model for Multi-Loci Genetic Mapping

In genome-wide association studies (GWAS), efficient incorporation of linkage disequilibria (LD) among densely typed genetic variants into analysis is a critical yet challenging problem. Functional linear models (FLM), which impose smoothing structure on the coefficients correlated covariates, are advantageous in mapping multiple with high LD. Here we propose novel constrained generalized FLM (cGFLM) framework to perform simultaneous tests block linked SNPs various trait types, including continuous, binary and zero-inflated count phenotypes. The new cGFLM applies set inequality constraints ensure model identifiability under different codings. method implemented via B-splines, an augmented Lagrangian algorithm employed for parameter estimation. For hypotheses testing, test statistic that accounts was derived, following mixture chi-square distributions. Simulation results show effective identifying causal loci gene clusters compared several competing methods based single markers SKAT-C. We applied proposed analyze candidate gene-based COGEND study large-scale GWAS data dental caries risk.