Multiple machine learning methods aided virtual screening of Na_V1.5 inhibitors

Abstract

Na v 1.5 sodium channels contribute to the generation of the rapid upstroke of the myocardial action potential and thereby play a central role in the excitability of myocardial cells. At present, the patch clamp method is the gold standard for ion channel inhibitor screening. However, this method has disadvantages such as high technical difficulty, high cost and low speed. In this study, novel machine learning models to screen chemical blockers were developed to overcome the above shortage. The data from the ChEMBL Database were employed to establish the machine learning models. Firstly, six molecular fingerprints together with five machine learning algorithms were used to develop 30 classification models to predict effective inhibitors. A validation and a test set were used to evaluate the performance of the models. Subsequently, the privileged substructures tightly associated with the inhibition of the Na v 1.5 ion channel were extracted using the bioalerts Python package. In the validation set, the RF‐Graph model performed best. Similarly, RF‐Graph produced the best result in the test set in which the Prediction Accuracy (Q) was 0.9309 and Matthew's correlation coefficient was 0.8627, further indicating the model had high classification ability. The results of the privileged substructures indicated Sulfa structures and fragments with large Steric hindrance tend to block Na v 1.5. In the unsupervised learning task of identifying sulfa drugs, MACCS and Graph fingerprints had good results. In summary, effective machine learning models have been constructed which help to screen potential inhibitors of the Na v 1.5 ion channel and key privileged substructures with high affinity were also extracted.

Dynamic Changes in Ion Channels during Myocardial Infarction and Therapeutic Challenges

“…Ion channels represent a large membrane protein superfamily critical for physiological functions in both excitable and non-excitable cells. While patch-clamp electrophysiology remains the gold standard for screening ion channel inhibitors, this approach is associated with technical challenges, high costs, and slow throughput [ 165 ]. With the advancement of cryo-electron microscopy (cryo-EM) technology, the structures of ion channels are becoming increasingly elucidated.…”

Recent Advances in Computer-Aided Structure-Based Drug Design on Ion Channels

“…The ChEMBL database includes currently data for ~2.3 million ligands and ~15,000 targets, which were collected from published assays, scientific publications, or gathered from other databases and research institutions across the world [102,103]. Screening experiments on ion channels often exploit the ChEMBL ion channel dataset [104,105]. One can specify a particular ion channel family and subtype and obtain a list of small molecule ligands that are known to act on these targets.…”

“…Mostly ligand-based virtual screening approaches are aided with machine learning. For example, Kong et al [105] designed multiple machine learning frameworks based on ligand molecular fingerprints to screen the ChEMBL database to identify inhibitors of Na V 1.5 as potential anti-arrhythmic drugs.…”

Machine learning study of the extended drug–target interaction network informed by pain related voltage-gated sodium channels

“…2 sodium channel inhibitors. 1,2,6,17,22,29,30,35,57 From 2018, more studies on VGSCs can be found in review articles. 24,37,40,43 However, these studies lack consideration of drug-target interaction networks, as well as comprehensive ADMET (absorption, distribution, metabolism, excretion, and toxicity) analysis.…”