We propose a computational framework for spatial niche analysis (SpaNiche) in spatial transcriptomics data to uncover colocalization patterns and infer potential ligand-receptor interactions. SpaNiche leverages graph-regularized joint non-negative matrix factorization to integrate information from cell abundance and ligand-receptor expression, identifying spatial colocalization patterns among cell types while providing insights into associated ligand-receptor interactions. Besides, SpaNiche employs consensus clustering to define "ecotypes", enhancing its utility in multi-sample spatial transcriptomics datasets. We investigate microenvironmental ecotypes in colorectal cancer, prostate cancer and cerebral cortex in early Alzheimer's disease, highlighting the broad applicability in dissecting complex colocalization patterns within spatial tissue microenvironments.
Huang et al. (Tue,) studied this question.