TABLE 1.

Application of AI in TCM discovery.

Aim of study	AI methods	Results	REF
Screen and validate the potential inhibitors of diabetes mellitus from TCM Database	Machine learning regression models, and deep neural network models	Illustrated that Hypecoum leptocarpum [Papaveraceae; Hypecoum leptocarpum Hook.f. & Thomson] might be a potential and effective medicine herb for the treatment of DM	Gong et al. (2021)
Identify potential inhibitor against phospholipase A2 (PLA2)	Virtual screening and molecular docking	Confirmed that Scutellarin [Lamiaceae; Scutellaria baicalensis Georgi] (TCM3290) had a potent snake venom neutralizing capacity	Chinnasamy et al. (2020)
Discover the molecular mechanism of Qu’s formula in treating endometriosis and to explore the potential drug targets	Network pharmacology, molecular docking	Confirm the clinical effectiveness of QUF	Wu et al. (2022)
Discover multi-target compounds with anti-inflammation activity from TCM	TCM target-effect relationship spectrum, virtual screening model	Seven compounds were chosen with multiple targeted activity based on the TCM-TERS	Huo et al. (2022)
Explore effective TCM against porcine reproductive and respiratory syndrome (PRRS)	Computer-aided drug design technology, virtual screening	Suggested that Andrographolide [Acanthaceae; Andrographis paniculata (Burm.f.) Nees] and Potassium Dehydrographolide Succinate are promising PRRSV inhibitors in vitro	Su et al. (2021)
Explore an integrated new approach for finding lead compounds that inhibit galectin-3	Deep learning-based algorithm	Found that the active ingredients 1,2-Dimethylbenzene and Typhic acid contained in Crataegus pinnatifida Bunge [Rosaceae; Crataegus Fructus] and Typha angustata [ Typhaceae; Typha domingensis Pers.] might be the effective inhibitors of neurodegenerative diseases	Deng et al. (2020)
Explore new targets and new active components of TCM in kidney stones	Molecular docking	Four active ingredients were predicted to have effects on treating kidney stones	Zhang et al. (2023)
Screen a drug targeting thioredoxin reductase (TrxR) from TCM database	Molecular docking	Identified two compounds as structure-based lead inhibitors against TrxR	Lata and Akif (2021)
Explore the possible target pathways for Gentianine [Gentianaceae; Gentiana scabra Bunge] in anti-ischemic stroke	Machine learning, molecular docking	Gentianine [Gentianaceae; Gentiana scabra Bunge] could be used as a potential anti-ischemic stroke agent by suppressing inflammatory responses via TLR4/NF-κB signaling pathway	Wang et al. (2021a)