A 3D Genome Atlas of Genetic Variants and Their Pathological Effects in Cancer.

The hierarchical organization of the eukaryotic genome is crucial for nuclear activities and cellular development. Genetic aberrations can disrupt this 3D genomic architecture, potentially driving oncogenesis. However, current research often lacks a comprehensive perspective, focusing on specific mutation types and singular 3D structural levels. Here, pathological changes from chromosomes to nucleotides are systematically cataloged, including 10 789 interchromosomal translocations (ICTs), 18 863 structural variants (SVs), and 162 769 single nucleotide polymorphisms (SNPs). The multilayered analysis reveals that fewer than 10% of ICTs disrupt territories via potent 3D interactions, and only a minimal fraction of SVs disrupt compartments or intersect topologically associated domain structures, yet these events significantly influence gene expression. Pathogenic SNPs typically show reduced interactions within the 3D genomic space. To investigate the effects of variants in the context of 3D organization, a two-phase scoring algorithm, 3DFunc, is developed to evaluate the pathogenicity of variant-gene pairs in cancer. Using 3DFunc, IGHV3-23's critical role in chronic lymphocytic leukemia is identified and it is found that three pathological SNPs (rs6605578, rs7814783, rs2738144) interact with DEFA3. Additionally, 3DGAtlas is introduced, which provides a highly accessible 3D genome atlas and a valuable resource for exploring the pathological effects of genetic mutations in cancer.