Multi-channel PINN: investigating scalable and transferable neural networks for drug discovery

Abstract

Analysis of compound–protein interactions (CPIs) has become a crucial prerequisite for drug discovery and drug repositioning. In vitro experiments are commonly used in identifying CPIs, but it is not feasible to discover the molecular and proteomic space only through experimental approaches. Machine learning’s advances in predicting CPIs have made significant contributions to drug discovery. Deep neural networks (DNNs), which have recently been applied to predict CPIs, performed better than other shallow classifiers. However, such techniques commonly require a considerable volume of dense data for each training target. Although the number of publicly available CPI data has grown rapidly, public data is still sparse and has a large number of measurement errors. In this paper, we propose a novel method, Multi-channel PINN , to fully utilize sparse data in terms of representation learning. With representation learning, Multi-channel PINN can utilize three approaches of DNNs which are a classifier, a feature extractor, and an end-to-end learner. Multi-channel PINN can be fed with both low and high levels of representations and incorporates each of them by utilizing all approaches within a single model. To fully utilize sparse public data, we additionally explore the potential of transferring representations from training tasks to test tasks. As a proof of concept, Multi-channel PINN was evaluated on fifteen combinations of feature pairs to investigate how they affect the performance in terms of highest performance, initial performance, and convergence speed. The experimental results obtained indicate that the multi-channel models using protein features performed better than single-channel models or multi-channel models using compound features. Therefore, Multi-channel PINN can be advantageous when used with appropriate representations. Additionally, we pretrained models on a training task then finetuned them on a test task to figure out whether Multi-channel PINN can capture general representations for compounds and proteins. We found that there were significant differences in performance between pretrained models and non-pretrained models. Electronic supplementary material The online version of this article (10.1186/s13321-019-0368-1) contains supplementary material, which is available to authorized users.

A review on machine learning approaches and trends in drug discovery

“…Extended Connectivity Fingerprints (ECFP) are a class of topological fingerprints for molecular characterization [26] , [27] . Historically, topological fingerprints were developed for the search of substructures and similarities, but these have been developed specifically for structure–activity modeling [28] , [29] , [30] , [31] , [32] .…”

“…In vitro experiments are commonly used to identify CPIs, but it is not feasible to perform this task through experimental approaches alone, and advances in machine learning in predicting CPIs have made great contributions to drug discovery. To improve the task of predicting ligand–protein interactions, tools such as Multi-channel PINN [26] . In addition, the study of these interactions is essential for obtaining traces of novel drugs and predicting their side effects from approved drugs and candidates [40] .…”

Artificial intelligence and machine learning‐aided drug discovery in central nervous system diseases: State‐of‐the‐arts and future directions

“…Various specially designed AI/ML models have been proposed for detecting novel drug indications. Here, we classify the ML applications for drug repositioning into the following three categories: (i) Similarity‐based methods that employ different types of classifiers like logistic regression, 305,306 SVM, 307–309 RF, 310,311 KNN, 312 and CNN, 313 (ii) feature vector‐based methods that utilize supervised 314–318 and semisupervised 319–321 learning algorithms, and (iii) network‐based methods that mainly use semisupervised learning algorithms (e.g., Laplacian regularized least square, 322–324 label propagation, 325 random walk, 326 and RF 310 ). We provide an in‐depth discussion of these three classes of AI‐based drug repositioning applications in the Supporting Information.…”

Artificial Intelligence in Drug Design: Algorithms, Applications, Challenges and Ethics

“…In another study, compound-protein interactions were targeted [128]. In this study, multichannel pairwise input neural networks were developed to benefit from sparse data in the literature.…”