Predicting Selective Serotonin Re-Uptake Inhibitors Potency: Machine Learning and Molecular Docking Approach

Selective Serotonin Re-uptake Inhibitors are medications employed for the treatment of mental diseases such as depression, anxiety disorders, and OCD. They enhance neurotransmission and mood control by increasing serotonin levels in the brain. This study employs ensemble machine learning and molecular docking to identify novel SSRIs. The potential ligands were obtained from CHEMBL database, which were then used to build a machine-learning model predicting the inhibitory constant (Ki) values of SSRIs. Molecular fingerprints served as feature descriptors in the model, which was trained using various ensemble machine-learning algorithms. The ExtraTrees Regressor with KlekothaRothFingerPrint as molecular fingerprint achieved the highest accuracy of 92% r-squared and 0.01 RMSE values. The ZINC database was queried to identify novel SSRIs, and the compounds with favorable Ki values underwent molecular docking simulations with the target molecule 7lwd. The study identified ZINC00427761, ZINC00427746, ZINC00425581, and ZINC00427764 as potential SERT inhibitors with strong molecular interactions and docking binding energies of -10.9, -10.8, 10.6, -10.1 kcal/mol respectively. These findings contribute to drug discovery efforts, particularly in the development of novel SSRIs. Additional in vitro and in vivo research are advised.