Multimodal 3D Genome Pre-training

Hui Xiong (Hong Kong University of Science and Technology (Guangzhou)) · Minghao Yang (The Hong Kong University of Science and Technology (Guangzhou Campus)) · Pengteng Li (HKUST(GZ)) · Yan Liang (South China Normal University) · Qianyi Cai (The Hong Kong University of Science and Technology) · Zhihang Zheng (Hong Kong Polytechnic University) · Shichen Zhang (Westlake University) · Pengfei ZHANG (The Hong Kong University of Science and Technology) · Zhi-An Huang (City University of Hong Kong (Dongguan))

3d genome structureaggregation cross-modal interaction downstream tasksepigenomic tracksfunctional implicationsheterogeneous semantic fusionhi-c contact mapslarge-scale datasetmapping blocksmix-hicmultimodal foundation modelpre-training state-of-the-art methodsunified representation

Deep learning techniques have driven significant progress in various analytical tasks within 3D genomics in computational biology. However, a holistic understanding of 3D genomics knowledge remains underexplored. Here, we propose ***MIX-HIC***, the first multimodal foundation model of 3D genome that integrates both 3D genome structure and epigenomic tracks, which obtains unified and comprehensive semantics. For accurate heterogeneous semantic fusion, we design the cross-modal interaction and mapping blocks for robust unified representation, yielding the accurate aggregation of 3D genome knowledge. Besides, we introduce the first large-scale dataset comprising over ***1 million*** pairwise samples of Hi-C contact maps and epigenomic tracks for high-quality pre-training, enabling the exploration of functional implications in 3D genomics. Extensive experiments show that MIX-HIC significantly surpasses existing state-of-the-art methods in diverse downstream tasks. This work provides a valuable resource for advancing 3D genomics research.