MetaLigand provides a prior-knowledge-guided framework for predicting non-peptide ligand mediated cell-cell communication

作者：Ying Xin, Yang Jin, Cheng Qian, Seth Blackshaw, Jiang Qian · 发表于：Cell Reports Methods · 年份：2025 · DOI：10.1016/j.crmeth.2025.101217 · 被引用次数：2 · 研究领域：Single-cell and spatial transcriptomics、Machine Learning in Bioinformatics、Bioinformatics and Genomic Networks

Non-peptide ligands (NPLs), including lipids, amino acids, carbohydrates, and non-peptide neurotransmitters and hormones, play a critical role in ligand-receptor-mediated cell-cell communication, driving diverse physiological and pathological processes. To facilitate the study of NPL-dependent intercellular interactions, we introduce MetaLigand, a tool designed to infer NPL availability and NPL-receptor interactions using transcriptomic data. MetaLigand compiles data for 233 NPLs, including their biosynthetic enzymes, transporter genes, and receptor genes, through a combination of automated pipelines and manual curation from comprehensive databases. The tool integrates both de novo and salvage synthesis pathways, incorporating multiple biosynthetic steps and transport mechanisms. Comparisons with existing tools demonstrate MetaLigand's ability to account for complex biogenesis pathways and model NPL availability across diverse tissues and cell types. Furthermore, analysis of single-nucleus RNA sequencing (RNA-seq) datasets from age-related macular degeneration samples revealed that distinct retinal cell types exhibit unique NPL profiles and participate in specific NPL-mediated pathological cell-cell interactions.