Haikuo Li

🧬 Spatial biology is moving beyond the transcriptome. A timely JCI Insight review maps the rapidly expanding landscape of spatial multiomics—technologies designed to measure not only where genes are expressed, but also where genomic alterations, epigenetic states, proteins, metabolites, and functional phenotypes occur within intact tissues. Spatial transcriptomics has transformed biomedical research, but RNA alone provides an incomplete picture. mRNA abundance may correlate poorly with protein levels or epigenetic states and cannot directly capture metabolic activity or small-molecule distributions. The next generation therefore asks: Can we reconstruct multiple layers of biology while preserving tissue architecture? The review organizes this rapidly evolving field into several major dimensions: 🧬 Spatial genomics Methods such as Slide-DNA-seq can map genomic alterations directly within tissues, allowing investigators to connect tumor clones with their spatial niches and distinguish genetically driven from microenvironment-dependent transcriptional programs. 🔐 Spatial epigenomics Spatial-ATAC-seq and spatial-CUT&Tag reveal chromatin accessibility and histone modifications in situ. Newer approaches can simultaneously capture epigenome + transcriptome + proteins, moving toward genuine spatial tri-omics. 🧪 Spatial proteomics Sequencing-based, imaging-based and mass-spectrometry platforms provide information that RNA cannot: protein abundance, localization, posttranslational modifications and potentially protein-protein interactions within disease niches. ⚡ Spatial metabolomics MALDI-MSI, DESI-MSI and label-free optical approaches map metabolites, lipids, redox states and even drug distributions—providing a more direct readout of cellular function and pharmacology. One particularly exciting frontier is spatial functional genomics: combining CRISPR perturbation with spatial readouts to determine how specific genetic changes reorganize neighboring cells and tissue ecosystems. Platforms including Perturb-map, Perturb-Multi and PERTURB-CAST are beginning to turn spatial atlases from descriptive maps into causal experiments. And AI may dramatically accelerate clinical translation. Instead of experimentally measuring every molecular layer, emerging models attempt to infer spatial molecular states directly from routine H&E histopathology, potentially creating scalable “virtual spatial profiling.” The field still faces major challenges: cost, resolution-versus-coverage tradeoffs, cell segmentation, cross-platform normalization, multimodal alignment and standardized benchmarking. But the trajectory is clear: Spatial transcriptomics → spatial multiomics → spatial functional biology → predictive tissue models. The ultimate goal is no longer simply to build increasingly detailed cell atlases. It is to understand how genome, epigenome, transcriptome, proteome and metabolome interact in space to create disease—and therapeutic response. 📄 Mao X, Chen Z, Hwang EJ, et al. Spatial multiomics in biomedical research: advances beyond transcriptomics. JCI Insight. 2026;11(17):e206951 🔗 DOI: 10.1172/jci.insight.206951 #SpatialBiology #SpatialOmics #Multiomics #SpatialTranscriptomics #Proteomics #Metabolomics #Epigenomics #SingleCell #PrecisionMedicine #AI

Excited to see this in print! Kudos to Haikuo, Bo, and the entire team including @MinaXu7 @yalepathology @zhou_jingtian @arcinstitute to explore this clinically relevant area of lymphoma transformation with truly innovative technology. @theNCI #HTAN nature.com/articles/s41467-0…