Deep Action: An approach on the basis of Deep Learning for the Prediction of Novel Drug-Target Interactions

In the processes of drug development and discovery, Drug-target interactions (DTIs) take part a vital position. DTI prediction through laboratory experiments consumes a lot of time. Also they were costly and tiring. Although computational approaches can recognize new interactions between drug-target pairs and speed up the drug conversion procedures, some problems like large scope of data and imbalanced class have been encountered in the course of the prediction procedures, and the number of unknown interactions were huge. Therefore, an approach on the grounds of deep learning (deepACTION) is put forward to predict possible or unrevealed DTIs. Here, each drug chemical structure and protein sequence is transformed according to structural and sequence information using different descriptors to correctly constitute their properties. In this method the majority and minority instances in the dataset are balanced using the SMOTE technique. For accurate DTI prediction a convolutional neural network (CNN) algorithm is trained with balanced and reduced features. For comparing the performance of the DeepACTION model with that of other methods AUC is regarded as the primary evaluation metric. An AUC curve of 0.933 is achieved by Deep ACTION model for the experimental dataset acquired from the Drug Bank database. Based on exper-imental results it is evident that the model is capable to predict a remarkable number of new DTI’s and it produce thorough knowledge that inspires scientists to instigate advanced drugs.