Exploring the Anti-Diabetic Potential of Indonesian Traditional Herbal Plants with Network Pharmacology Approach

Diabetes Mellitus (DM) is a chronic metabolic disease with increasing prevalence, and it remains a major health problem worldwide. This study investigated the anti-diabetic potential of Indonesian traditional herbal plants ( Curcuma amada Roxb., Curcuma longa , and Allium cepa L. var. aggregatum ) using a network pharmacology approach combined with molecular docking. Active compounds were collected from IJAH Analytics and KNApSAcK, screened using oral bioavailability and drug-likeness criteria, and then mapped to predicted protein targets using SwissTargetPrediction. Diabetes-related targets were obtained from OMIM, UniProt, MalaCards, and GeneCards, and overlapping targets were selected for network analysis. A protein--protein interaction network was built using STRING v12.0, followed by centrality analysis (degree, betweenness, closeness, and eigenvector). The Skyline query algorithm was applied to prioritize key proteins, and the top five proteins were selected as receptors for docking. Molecular docking was performed using AutoDock Vina against nine candidate ligands. The docking results showed that Tropeoside B1 consistently produced the strongest and most frequent binding across the selected receptors, followed by Progesterone and Peonidin-3-Arabinoside. Overall, this workflow helps narrow down promising herbal compounds and protein targets, providing candidates for further experimental validation.