Prediction and Optimization of Na_V1.7 Sodium Channel Inhibitors Based on Machine Learning and Simulated Annealing

Abstract

Although the NaV1.7 sodium channel is a promising drug target for pain, traditional screening strategies for discovery of NaV1.7 inhibitors are very painstaking and time-consuming. Herein, we aimed to build machine learning models for screening and design of potent and effective NaV1.7 sodium channel inhibitors. We customized the imbalanced data set from ChEMBL and BindingDB to train and filter the best classification model. Then, the whole-cell voltage-clamp was employed to validate the inhibitors. We assembled a molecular group optimization method by combining the Grammar Variational Autoencoder, classification model, and simulated annealing. We found that the RF-CDK model (random forest + CDK fingerprint) performs best in the imbalanced data set. Of the three compounds that may have inhibitory effects, nortriptyline has been experimentally verified. In the molecule optimization process, 40 molecules located in the applicability domain of RF-CDK were used as a starting point, among which 34 molecules evolved to molecules with greater molecular scores (MS). The molecule with the highest MS was derived from CHEMBL­2325245. The model and method we developed for NaV1.7 inhibitors are also applicable to other targets.

Multiple machine learning methods aided virtual screening of NaV1.5 inhibitors

“…The average similarity value of the whole positive molecule set before undersampling samples is 0.301 (Figure 2A); the average similarity value after undersampling s is 0.238 (Figure 2B) and the average similarity of the whole negative molecule set was only 2C). And compared with other research results, [39][40][41] the molecule similarity here (0.238 for positive molecules and 0.204 for negative molecules) has lower values. The above results showed that clustering and undersampling can effectively reduce molecular similarity in the positive molecules and there is no need to do clustering and undersampling in the negative molecules.…”

Simulation and Machine Learning Methods for Ion-Channel Structure Determination, Mechanistic Studies and Drug Design

“…Based on the data collected from ChEMBL, BindingDB, and in-house databases, various machine learning-based QSAR models were built by Kristam et al to predict blockers of the voltage-gated sodium ion channel Na v 1.5, with the balanced accuracy of 0.88 (at the threshold of 1 μM) and predicted R 2 of 0.71 (RMSE = 0.73 for pIC50) for the classification and regression models, respectively ( Khalifa et al, 2020 ). Similarly, Huang and Xie et al trained and filtered a classification model based on the data from ChEMBL and BindingDB for the discovery of Na v 1.7 blockers ( Kong et al, 2020 ). The Grammar Variational Autoencoder, the trained classification model, and simulated annealing were combined to conduct the molecular optimization and an active compound was found and identified experimentally ( Kong et al, 2020 ).…”

“…Similarly, Huang and Xie et al trained and filtered a classification model based on the data from ChEMBL and BindingDB for the discovery of Na v 1.7 blockers ( Kong et al, 2020 ). The Grammar Variational Autoencoder, the trained classification model, and simulated annealing were combined to conduct the molecular optimization and an active compound was found and identified experimentally ( Kong et al, 2020 ). Li et al studied the pharmacophore hypothesis of ligands that bind at the benzodiazepine site of GABA A receptors based on ligand-based pharmacophore, 3D-QSAR analysis, and Bayesian models, which might provide useful viewpoints in the discovery of GABA A modulators ( Yang Y. et al, 2013 ).…”

Artificial intelligence and machine learning‐aided drug discovery in central nervous system diseases: State‐of‐the‐arts and future directions

“…The other clinical applications of AI systems in pain management include prediction of pain severity/modality and analgesic requirements, 441–443 individualized medicine decision support in analgesic treatment, 444,445 prediction of the effectiveness of the analgesics, 446,447 and prediction of medication overuse 448–450 . Besides the clinical applications, researchers have employed ML methods at the early stages of analgesic discovery, such as identifying novel genes and pathways associated with acute and chronic pain 451 and predicting inhibitors of a drug target for pain (i.e., NaV1.7 sodium channel) 452 . To facilitate the prediction of novel multi‐target analgesics or drug combinations for pain treatment, researchers have established a comprehensive pain‐domain‐specific chemogenomics knowledgebase that includes the analgesics in current use, pain‐related targets with all available 3D structures, and the compounds reported for these target proteins 453 …”