Detecting QTL Hotspots in Sparse Regression Models

This lecture by the instructor focuses on large-scale inference for detecting QTL hotspots in hierarchically-related sparse regression models. The presentation covers the transition from GWAS to functional genomics, emphasizing the need to use genomics to understand the variation in phenotypes, disease susceptibility, and drug responses. The lecture introduces molecular QTL studies, highlighting the differences from classical GWAS and the challenges in statistical modeling due to high-dimensional data. The fully joint QTL mapping framework is discussed, emphasizing the use of hierarchical models to detect 'hotspot SNPs' controlling multiple traits. Variational inference methods and the impact of supplying informative epigenetic marks are also explored.