Cell3D: a comprehensive database for interactive 3D visualization and crosstalk analysis of single-cell and spatial omics
Qian Liu, Yuting He, Rui Sun, Baokui Wang, Ran Cui, Xinxin Shan, Borui Xu, Zhehe Xu, Ximing Zhang, Yue Gao, Shangwei Ning, Li Wang, Peng Wang, Qiuyan GuoAbstract
Single-cell and spatial transcriptomics yield large-scale, high-dimensional datasets. However, cross-feature integrative exploration is constrained, since most current platforms support a single data type and lack interactive visualization for multi-feature analysis. To address this unmet need, we developed Cell3D, an accessible and versatile interactive platform designed to advance integrative multi-level feature analysis. Cell3D curates 898 publicly available datasets, comprising 397 single-cell and 501 spatial transcriptomic datasets, encompassing 2 924 646 single cells and spatial spots across 105 diseases, 756 cell types, and 83 organs/tissues from 11 species. The database integrates twelve mainstream functional annotation systems, providing a comprehensive landscape for dissecting molecular features across spatial and cellular dimensions. Users can explore data at multiple levels, from gene expression and cell clusters to pathway activities and clinical phenotypes. Cell3D offers a suite of user-friendly functionalities, including customizable 2D/3D interactive visualization for dynamic multi-level feature projection, flexible data integration with automatic recognition, quantitative crosstalk analysis, and microenvironment-focused integrative analysis supporting disease research and biomarker discovery. Collectively, Cell3D enables exploration of tissue microenvironments at both cellular and spatial resolutions, bridging molecular perturbations and tissue architecture in health and disease. The database is freely accessible at https://bio-bigdata.hrbmu.edu.cn/Cell3D.