A Cross‐Species Hormone Knowledge Base With Multi‐Dimensional Data Integration and Intelligent Query Assistance

Abstract

Multi‐source biological data have provided a solid foundation for deciphering hormone regulatory networks, yet systematic frameworks that intelligently integrate and exploit such resources are scarce. Here, we present AHormoneDB ( https://panhorm.cn ), a database incorporating multi‐layer omics and a novel pipeline for hormone‐target gene prediction. Our core algorithm, GeneReg v2.6, employs a weighted evidence‐scoring system integrating ChIP‐seq, conserved motifs, and literature mining. In an internal test set, it achieved an F1‐score of 0.894 (95% CI 0.818–0.955). To facilitate downstream interpretation, AHormoneDB harmonizes literature‐mined hormone–gene–disease associations with disease ontology (DOID) standardization and embeds a DeepSeek‐based retrieval‐augmented generation assistant for exploratory, natural‐language querying to complement structured data retrieval. In an acromegaly case study, AHormoneDB both recapitulated JAK–STAT/PI3K–AKT cascades and nominated PTPN1 as a previously underappreciated candidate node for further investigation. The platform offers a no‐code interface for regulatory network visualization and drug‐target mining. All source code is publicly available on GitHub, and the database is accessible at https://panhorm.cn .