SMILES-Based QSAR and Molecular Docking Study of Oseltamivir Derivatives as Influenza Inhibitors

Abstract

The quantitative structure–activity relationship studies for the modeling the activity of 72 oseltamivir derivatives as influenza neuraminidase (H1N1) inhibitors are performed using the Monte Carlo method based on the target function involving index of ideality of correlation (IIC). The optimal descriptors based on the combination of SMILES and hydrogen suppressor graphs (HSG) are employed for the model construction. Internal and external validation confirms robustness and good predictive power of the generated QSAR models. Identification of the activity-enhancing attributes indicates the positive impact of nitrogen and double bond on the influenza inhibitory activity. Finally, the pIC50 of the twelve new oseltamivir derivatives from ChEMBL database were predicted based on the proposed model. The new compounds showed high predicted pIC50 values and their molecular docking study was also investigated.