Bridging the variant-to-function gap in type 2 diabetes: advances and challenges.

Type 2 diabetes is a polygenic, heterogeneous disease affecting over 530 million individuals worldwide, a number projected to rise to 1.3 billion by 2050. Genome-wide association studies have identified over 1200 loci associated with type 2 diabetes, yet only a modest proportion of disease heritability has been explained, and most risk variants lie in non-coding regions, complicating biological interpretation. This review synthesises how recent advances in variant-to-function (V2F) strategies, including fine-mapping, chromatin profiling, single-cell multi-omics, scalable CRISPR-based perturbation strategies and machine learning tools, have begun to link non-coding variants to effector genes, tissues and molecular mechanisms. We highlight examples of V2F resolution across key type 2 diabetes tissues, including the roles of TCF7L2 in pancreatic beta cell development, SLC16A11 and PNPLA3 in hepatocyte lipid metabolism, and GLUT4 trafficking pathways in adipocytes and muscle. We provide a summary table of genetically anchored, experimentally supported V2F discoveries in type 2 diabetes, and describe emerging tools and frameworks to address disease heterogeneity and context dependence. Together, these developments are on the cusp of closing the genetic association and mechanism gap while laying the foundation for future precision medicine approaches in type 2 diabetes.