Integrating Spatial Proteogenomics in Cancer Research

BACKGROUND: Spatial proteogenomics marks a paradigm shift in oncology by integrating molecular analysis with spatial information from both spatial proteomics and other data modalities (e.g., spatial transcriptomics), thereby unveiling tumor heterogeneity and dynamic changes in the microenvironment. METHODS: We systematically reviewed the evolution of spatial proteogenomics, from single-modality profiling to integration with transcriptomics and metabolomics, from the detection of abundant proteins to exploration of "dark proteome" with low abundance or stability, and from analytic software based on traditional machine learning algorithms to advanced artificial intelligence-driven analytical frameworks. RESULTS: Key advances of sub-fields of spatial proteogenomics include: RNA-protein co-localization: Spatial CITE-seq, enabling RNA-protein co-localization to reveal immune microenvironmental patterns and neoantigen distribution. Spatial Proteomics + Spatial Metabolomics: Matrix-assisted laser desorption/ionization imaging (MALDI), overcoming protein detection bottlenecks and capturing metabolic reprogramming. Deep visual proteomics (DVP): achieving unbiased spatial analysis via AI-guided microdissection. Spatial-aware multiplex dark proteome approaches: Examples are nanodroplet processing in one pot for trace samples (NanoPOTS) and proteoform imaging mass spectrometry (PiMS). Multimodal foundation AI models: Examples are KRONOS and HEIST, which integrate multiple data modalities and significantly improve diagnostic precision and therapeutic prediction. CONCLUSIONS AND FUTURE DIRECTIONS: Despite challenges of resolution, standardization, and data complexity, spatial proteomics is advancing rapidly. Together with frontier technologies such as quantum computing, live imaging, and organoid integration, it is driving breakthroughs in cancer diagnosis, personalized immunotherapy, and drug development.