MedMeta: An AI-Enabled and Genomics-Based Database for Functional Profiling of Secondary Metabolites in Medicinal Species

Abstract

Medicinal resources contain a vast array of secondary metabolites that play critical roles in disease treatment, health maintenance, and drug discovery. Nevertheless, challenges such as biosynthetic complexity and species-specific variability have long hindered their systematic exploration. Recent advances in omics technologies and artificial intelligence (AI)-driven approaches have opened new avenues via which to decode biosynthetic pathways and discover secondary metabolites using omics-level data. In this study, we present MedMeta, a curated and integrative database that connects secondary metabolites with genomic, biochemical, and pharmacological information across 1035 medicinal species documented in eight authoritative global pharmacopoeias. MedMeta comprises 146 101 predicted active secondary metabolites, 196 356 biosynthetic pathways, and an extensive set of annotated molecular targets. As a proof of principle, we employed MedMeta to investigate three representative Apiaceae species—Pucedanum praeruptorum, Angelica sinensis, and Apium graveolens—demonstrating its ability to uncover species-specific metabolite profiles, validate enzymatic functions, and identify compounds with important therapeutic potential. Overall, MedMeta can provide a powerful and scalable platform for natural product research, supporting both fundamental studies and applied biomedical applications. This database offers an invaluable resource for compound discovery, synthetic biology, geoherbalism studies, and the modern application of traditional medicinal systems.