DEEPScreen: High Performance Drug-Target Interaction Prediction with Convolutional Neural Networks Using 2-D Structural Compound Representations

Abstract

The identification of physical interactions between drug candidate chemical substances and target biomolecules is an important step in the process of drug discovery, where the standard procedure is the systematic screening of chemical compounds against pre-selected target proteins. However, experimental screening procedures are expensive and time consuming, therefore, it is not possible to carry out comprehensive tests. Within the last decade, computational approaches have been developed with the objective of aiding experimental studies by predicting novel drug-target interactions (DTI), via the construction and application of statistical models. In this study, we propose a large-scale DTI interaction prediction system, DEEPScreen, for early stage drug discovery, using convolutional deep neural networks. One of the main advantages of DEEPScreen is employing readily available simple 2-D images of compounds at the input level instead of engineered complex feature vectors that displayed limited performance in DTI prediction tasks previously. DEEPScreen learns complex features inherently from the 2-D molecular representations, thus producing highly accurate predictions.DEEPScreen system was trained for 704 target proteins (using ChEMBL curated bioactivity data) and finalized with rigorous hyper-parameter optimization tests. We compared the performance of DEEPScreen against shallow classifiers such as the random forest, logistic regression and support † A preliminary version of this study have been orally presented at vector machines, to indicate the effectiveness of the proposed deep learning approach. Additionally, we compared DEEPScreen with other deep learning based state-of-the-art DTI predictors on widely used benchmark datasets and showed that DEEPScreen produces better or comparable results to the top performers. The method proposed here can be employed to computationally scan a large portion of the recorded drug candidate compound and protein spaces to aid the experimentalists working in the field of drug discovery and repurposing by providing a preselection of interesting novel DTIs. of correctly predicted non-interacting compound-target pairs, whereas FP (i.e., false positive) represents the number of non-interacting compound target pairs, which are predicted as interacting.

MatchMaker: A Deep Learning Framework for Drug Synergy Prediction

“…This means accurate and efficient computational prioritization of novel synergistic drug pair candidates will keep being an important research area. In MatchMaker, we use only the drug chemical structure as the primary feature, which has been used in parallel tasks such as drug target identification [50] or drug side effect prediction [51,52]. We also use the cell line specific gene expression profile to capture the context of the experiment.…”

Using Single Protein/Ligand Binding Models to Predict Active Ligands for Unseen Proteins

“…tions, to simultaneously predict interactions between multiple protein targets and multiple candidate ligands or drugs (16)(17)(18)(19)(20). DTI models use the experimental data available for some subset of interactions ( Fig.…”

Learning Drug Functions from Chemical Structures with Convolutional Neural Networks and Random Forests

“…Finally, images and CNNs have been used to predict drug–protein interactions and outperformed models trained on flattened versions of the images, which cannot exploit the spatial structure. 31…”

“…These authors found that augmenting the same deep-learning architecture with only three additional chemical properties further improved model performance, suggesting that chemical images alone may not entirely capture the important characteristics of a chemical. Finally, images and CNNs have been used to predict drug–protein interactions and outperformed models trained on flattened versions of the images, which cannot exploit the spatial structure …”