Abstract
Spatial transcriptomics (ST) technologies enable gene expression profilingwith spatial resolution, offering unprecedented insights into tissueorganization and disease heterogeneity. However, current analysis methods oftenstruggle with noisy data, limited scalability, and inadequate modelling ofcomplex cellular relationships. We present SemanticST, a biologically informed,graph-based deep learning framework that models diverse cellular contextsthrough multi-semantic graph construction. SemanticST builds multiplecontext-specific graphs capturing spatial proximity, gene expressionsimilarity, and tissue domain structure, and learns disentangled embeddings foreach. These are fused using an attention-inspired strategy to yield a unified,biologically meaningful representation. A community-aware min-cut loss improvesrobustness over contrastive learning, particularly in sparse ST data.SemanticST supports mini-batch training, making it the first graph neuralnetwork scalable to large-scale datasets such as Xenium (500,000 cells).Benchmarking across four platforms (Visium, Slide-seq, Stereo-seq, Xenium) andmultiple human and mouse tissues shows consistent 20 percentage gains in ARI,NMI, and trajectory fidelity over DeepST, GraphST, and IRIS. In re-analysis ofbreast cancer Xenium data, SemanticST revealed rare and clinically significantniches, including triple receptor-positive clusters, spatially distinctDCIS-to-IDC transition zones, and FOXC2 tumour-associated myoepithelial cells,suggesting non-canonical EMT programs with stem-like features. SemanticST thusprovides a scalable, interpretable, and biologically grounded framework forspatial transcriptomics analysis, enabling robust discovery across tissue typesand diseases, and paving the way for spatially resolved tissue atlases andnext-generation precision medicine.