Drug Target Interaction Prediction Using Convolutional Neural Network (CNN)

Abstract

The goal of this research is to construct an artificial intelligence-based machine learning and deep learning system for predicting drug-target interactions. In this study, we categorize drug-target interactions across various drug combinations using a convolutional neural network (CNN) model. Deep learning-based CNN achieves state-of-the-art performance on the DDI-Corpus dataset, with an F1-score of 0.82 0.012 for the single model and 0.81 0.015 for the ensemble model, and a validation accuracy of 96.72 %, Our findings are intriguing, but it's important to remember that various types of scientific journals are essentially identical. When clinical reports and empirical data are compared, the discrepancies become increasingly evident. Furthermore, each category contains numerous outliers. Additionally, unusual pharmacological routes or DDIs are feasible. The inner workings of such a temporal system are beyond the scope of the current qualitative investigation but will be the focus of a future qualitative investigation. Future research aims to identify informational gaps in the existing literature that may lead to the discovery of concealed discovery indicators (DDIs).