SMILES-based QSAR and molecular docking studies of chalcone analogues as potential anti-colon cancer

Abstract

Abstract QSAR modeling was applied to predict the anti-colon activity (against HT-29) of 193 chalcone derivatives using the Monte Carlo method, based on the index of ideality correlation (IIC) target function. The models were constructed using CORAL software, which employed optimal descriptors combining SMILES notation and hydrogen-suppressed molecular graphs (HSG). Among the developed models, Split #2 was identified as the best-performing model, with R2_validation = 0.90, IIC_validation = 0.81, and Q2_validation = 0.89. The mechanistic interpretation of the models, utilizing enhancing/reducing promoters, demonstrated that the models are capable of accurately predicting the pIC50 values of other chalcone derivatives with high robustness and precision. Based on these promoters, ten new compounds were selected from the ChEMBL database for pIC50 prediction, and molecular docking was performed using the protein with PDB ID:1SA0.