Anti-Fungal Activity Prediction of Curcuma Longa and Piper belte through Molecular Docking

Abstract

Traditional medicine is deeply rooted in Indonesian culture, yet concerns persist regarding the safety and efficacy of medicinal plants. This study focuses on Curcuma longa and Piper betle, two prevalent medicinal plants in Southeast Asia, known for their potential antifungal properties, though the specific bioactive compounds remain unidentified. A machine learning model was developed using a dataset from DUD-E (Directory of Useful Decoys) docking to predict antifungal compounds, successfully identifying substances capable of inhibiting Cytochrome P450 EryK protein. Notably, compounds such as curcumin, beta-caryophyllene, and eugenol were highlighted for their significant antifungal activity. From the analysis, 21 compounds from Curcuma longa and 15 from Piper betle were selected and screened for their inhibitory potential. These compounds were further filtered based on Lipinski's rule of five, resulting in 7 compounds from Curcuma longa and 5 from Piper betle exhibiting binding affinities comparable to the reference ligand. Tanimoto similarity analysis revealed that the identified compounds shared less than 10% similarity in chemical structure to the reference ligand. However, the binding amino acids in several compounds demonstrated over 50% similarity to those of the reference ligand. The findings underscore the therapeutic potential of these compounds, contributing to the development of natural antifungal agents.