LSGI: interpretable spatial gradient analysis for spatial transcriptomics data

• Published in: Genome Biology Vol. 26; no. 1; pp. 238 - 19
• Main Authors: Liang, Qingnan, Solis Soto, Luisa, Haymaker, Cara, Chen, Ken
• Format: Journal Article
• Language: English
• Published: London BioMed Central 08.08.2025; BMC
• Subjects: Advances in Spatial Transcriptomics for Understanding Development and Disease; Animal Genetics and Genomics; Bioinformatics; Biomedical and Life Sciences; Cells; Computational Biology - methods; Datasets; Evolutionary Biology; Gene expression; Gene Expression Profiling - methods; Genotype & phenotype; Glioma; Human Genetics; Humans; Hypoxia; Investigations; Life Sciences; Method; Microbial Genetics and Genomics; Neoplasms - genetics; Neoplasms - metabolism; Plant Genetics and Genomics; Transcriptome; Transcriptomics; Tumors
• ISSN: 1474-760X; 1474-7596; 1474-760X
• DOI: 10.1186/s13059-025-03716-1

Cellular anatomy and signaling vary across niches, which can induce gradated gene expressions in subpopulations of cells. Such spatial transcriptomic gradient (STG) makes a significant source of intra-tumor heterogeneity. We present Local Spatial Gradient Inference (LSGI), a computational framework that systematically identifies spatial locations with prominent, interpretable STGs from spatial transcriptomic (ST) data. We demonstrate LSGI in tumor ST datasets and identify pan-cancer and tumor-type specific pathways with gradated patterns, highlighting the ones related to spatial transcriptional intratumoral heterogeneity. LSGI enables interpretable STG analysis, which can reveal novel insights in tumor biology from the increasingly reported tumor ST datasets.