Integrative Statistical Approaches for Complex Disease Study

Abstract

Advances in genomics and multi-omics technologies have opened new avenues for understanding complex traits and disease heterogeneity, both of which are critical for the development of precision medicine. Genome-wide association studies (GWASs) have identified numerous loci associated with complex traits, yet pinpointing causal variants and underlying biological mechanisms remains challenging. In this context, we will discuss a fine-mapping approach that integrates GWAS summary statistics with functional annotations to help prioritize likely causal variants. Alongside this, we will talk about an information-guided clustering framework designed for disease subtyping using high-dimensional multi-omics data. This method incorporates prior biological knowledge through a group lasso penalty, supporting both pattern discovery and feature selection across diverse data types. Together, these complementary strategies highlight the potential of statistical methods that combine heterogeneous data sources to refine genetic insights and improve our understanding of disease complexity.