AnnoTCM: An Annotation Tool for Multimodal Metabolite–Phenotype Integration in Decoding Bioactive Compounds of Traditional Chinese Medicine

Despite decades of pharmacological interest, the systematic identification of bioactive compounds in natural products (NPs) and traditional Chinese medicines (TCMs) remains constrained by the chemical complexity of herbal formulations and the absence of scalable tools to link chemical features with functional phenotypes. The field of drug discovery based on TCMs is a complex area of research that necessitates robust analytical platforms. While untargeted mass spectrometry enables the high-throughput profiling of molecular constituents, and high-content imaging captures multiparametric cellular responses, these data sets remain largely disconnected, limiting their translational utility. To address this gap, we developed AnnoTCM, an open-access platform integrating high-content cell phenotypic screening with untargeted metabolomics to systematically annotate pharmacologically active ions in TCMZ-score normalization and cluster correlation. To validate the utility of the platform, we employed Baoyuan Decoction (BYD), a classical antiaging traditional Chinese medicine formula. Using BYD, we constructed a database comprising 48 448 ions and associated phenotypic profiles derived from inflammation- and senescence-relevant models. Through the scoring and categorization of active ions, we identified 16 active compounds, including apigenin-7-O-glucoside and ginsenoside Rg1. These compounds were subsequently validated experimentally. Collectively, these findings demonstrate that AnnoTCM provides a generalizable, function-oriented platform for compound prioritization in TCMs research. This platform supports data-driven pharmacological studies and facilitates early-stage drug discovery from natural products.