Computational Decoding of Cell-Cell Communications in Heterogeneous Cellular Microenvironments Based on Spatial Transcriptomics.

Cell-cell communication (CCC) is a fundamental message-passing process through which biochemical signals help coordinate cell behaviors and functions. Understanding CCC is crucial to uncovering the driving force of cellular dynamics, as well as providing biological insights into immune response coordination and response to outer factors like virus infections. Studies of CCC have traditionally relied on bulk and single-cell transcriptomics. As the study of CCC progresses, researchers have a growing interest in capturing the spatial context, temporal dynamics, and heterogeneous interactions across multiple cell types. The advent of spatial transcriptomics (ST) technologies has reformed the field by enabling the high-resolution and high-coverage mapping of gene expression with annotated spatial coordinates, offering a great opportunity to dissect more accurate and biologically interpretable cellular interactions. At the same time, the development of advanced computational approaches, including statistical modeling, network analysis, and machine learning, has demonstrated the great potential to integrate multimodal spatial data and infer complex networks of cellular communications. In recent years, spatially resolved omics and computational modeling together are reshaping the field of CCC inference, leading to new investigations of more biologically meaningful problems. Here, we introduce the current landscape of CCC studies, emphasizing how ST data provide spatially resolved insights, and discuss the integration of novel algorithms to enhance CCC inference. We illustrate some representative methods and frameworks and point out extensions that leverage ST and AI-powered tools to advance the understanding of cellular communications.