Identification of Natural Product Inhibitors Targeting Dengue Capsid Protein Using an Open-Access Artificial Intelligence-Based Drug Discovery Methodology

Abstract

Dengue fever, or break-bone fever, is caused by dengue flavivirus transmitted through mosquito bites. To identify a druggable protein target for developing effective antiviral therapies, we studied how proteins from the dengue virus interact with the human body and created a detailed protein-protein interaction network of these interactions. The analysis of molecular functions and biological processes associated with proteins has shown that the capsid protein is crucial in facilitating the interaction between the virus and the host. This finding highlights the significance of the capsid protein as a potential target. We used AutoDockFR to study the binding of 208 natural compounds from Azadirachta indica with capsid protein. We screened the compounds using the X-ray diffraction structure of capsid protein (6vg5) and placed them into the binding pocket of an inhibitor called ST-148. We re-docked the inhibitor ST-148 and considered its docking score (-8.5) as the threshold value for hit selection. After applying these criteria, we obtained 81 hits. The binding mode analysis of the hits revealed that the cyclopentanoperhydrophenathrene ring structure is an essential pharmacophore that fits well into the binding cavity. Further molecular dynamics simulation study of the complexes for the highest affinity and the lowest affinity hits score confirmed the stability of the complex. An exhaustive analysis of the physical and chemical characteristics of potential drug candidates and their pharmacokinetic profiles has revealed that the natural products under consideration hold great potential as a viable treatment option for infections caused by the dengue virus.