Machine learning, molecular docking and simulation studies reveal Lomitapide, Lodipamide, Zafirlukast, Netupitant and Salmon Calcitonin as potent drug molecules against Chagas disease

Abstract

ABSTRACTTrypanosoma cruzi(T.cruzi), a protozoan parasite, is the pathogen that causes the hazardous disease Chagas, sometimes referred to as American trypanosomiasis. Currently, there are just two drugs commercially available to treat this fatal condition. So, there is an urgent requirement to create novel, safer and efficient anti-Tc drugs. In this study, we introduce a robust experimental strategy that integrates machine learning with molecular docking and simulation studies to identify the most potential drug candidates from the DrugBank dataset for treating the Chagas disease. In a machine learning method, different classifiers (Naïve Bayes, Random Forest, SMO, and C4.5) were used to train the model on the PubChem dataset and the most effective model (C4.5 algorithm) was then chosen and tested on the DrugBank dataset (containing FDA-approved and investigational drugs). The C4.5 algorithm-based machine learning model with an accuracy of 65% predicts the possible drug candidates (a total of 280 drugs were predicted). AutoDock4.2 software was then used to dock the predicted compounds that had a confidence interval of 80% and above. As a result, 47 predicted drugs were docked, and the best of the 5 drugs based on their docking score were chosen for performing the MD simulation studies. MD simulations (100 ns) were conducted for each of the protein-ligand complexes, which produces an average RMSD score (2.23 Å). The RMSF value of the cruzain (Cys25) binding pocket, after interaction with the five compounds were found to be less than 3.0 Å. Therefore, based on the molecular docking and simulation studies, we found out that Lomitapide, Lodipamide, Zafirlukast, Netupitant and Salmon Calcitonin have shown a stronger binding affinity with the crystal structure of cruzain (Cyst25) molecule, making them potential effective inhibitors. Hence, we could anticipate that our five computationally validated drugs will soon be available at the clinical trial stage and eventually be made accessible to the necessary Chagas disease patients.