Classification Of Human Ether-a-go-go Related Gene Chanel Blocking Compounds

Inhibition of the human ether-a-go-go related gene (hERG) potassium ion channel is one of the major factors related to severe cardiotoxicity leading to long QT syndrome (LQTS), and it is a predisposing factor for syncope and sudden death. The commonly used assay methods for this purpose include traditional patch clamp techniques, radioligand binding assays, cell-based fluorescence assays and Rb+-flux assays. However, for early-stage safety assessment, those experimental methods often lack throughput and are too expensive. So it is necessary to in silico tools to filter out potential hERG channel inhibitors in early stages of the drug discovery process, and it can improve the efficiency of drug discovery. This thesis studied on hERG blockage.Here, we describe several binary classification models based on a large and diverse dataset library of1969compounds. The10μM was used as the threshold for hERG channel blockade. MOE and MACCS Fingerprint descriptors were used to produce the models. The10-fold cross-validation method was used.10%of both hERG blockers and non-hERG blockers were randomly selected from the datasets and separated from the training set to be used later as an external test set. The model based on MACCS Fingerprint showed the best performance. It achieved an average overall accuracy up to90%with an average Matthews coefficient correlation (MCC) of0.766. In order to validate the stability of our model, we used y-scrambling on the best obtained model. The results of the y-scrambling proved the model to be efficient.