Abstract
MOTIVATION: Family-based study designs enable genetic association analyses that are robust against population stratification and admixture. The Family-Based Association Test (FBAT) framework extended and generalized the Transmission Disequilibrium Test concept for trios (affected offspring and their parents) to general pedigrees, arbitrary phenotypes, and multi-variant analyses. Following the successful identification of disease susceptibility loci in classical family studies, FBAT approaches can be a valuable tool in modern large-scale biobanks that contain significant proportions of related individuals, often with heterogeneous genetic ancestries. However, recent methods for established downstream analyses, such as statistical fine-mapping, are primarily tailored towards genetic association results obtained from regression models in unrelated individuals, and suitable methodologies for FBAT results have not been established in the literature. Here, we introduce and implement a framework for statistical fine-mapping in FBAT-based family-based association studies. RESULTS: Our approach is based on the established Deterministic Approximation of Posteriors algorithm and utilizes the association z-scores obtained from FBAT. We illustrate the method in simulation studies and an application to the Apolipoprotein E locus in a family-based association study of Alzheimer's Disease. AVAILABILITY AND IMPLEMENTATION: The fine-mapping approach FBAT DAP-G is implemented in the FBAT package https://github.com/FBATsw/FBAT/. The original code of DAP-G is available at https://github.com/xqwen/dap.