Intercellular Communication Guides the Prediction of Intracellular Gene Regulatory Relationships

Cellular communication via ligand-receptor signaling regulates downstream gene expression networks, playing a vital role in maintaining cellular functions and driving disease progression. However, the current methods do not account for the synergistic interactions between cellular communication and downstream gene regulatory networks. Moreover, existing approaches cannot construct complete cellular communication networks, thereby limiting biological significance and interpretability. To address these gaps, we propose a computational framework that predicts intracellular gene regulatory relationships by constructing comprehensive cellular communication networks. The framework introduces two key innovations: (1) end-to-end modeling from extracellular signals to gene expression by integrating ligand-receptor interactions, signaling pathway activation, and transcription factor regulatory networks and (2) accurate modeling of receptor-mediated regulatory relationships using deep learning to reveal intracellular mechanisms driven by cellular communication. Experimental results and case studies show that the framework efficiently predicts receptor-mediated target gene regulatory relationships across diverse spatial transcriptomic data sets and provides a valuable tool for uncovering biological processes.