Abstract:BackgroundIn recent years, research in artificial neural networks has resurged, now under the deep-learning umbrella, and grown extremely popular. Recently reported success of DL techniques in crowd-sourced QSAR and predictive toxicology competitions has showcased these methods as powerful tools in drug-discovery and toxicology research. The aim of this work was dual, first large number of hyper-parameter configurations were explored to investigate how they affect the performance of DNNs and could act as starting points when tuning DNNs and second their performance was compared to popular methods widely employed in the field of cheminformatics namely Naïve Bayes, k-nearest neighbor, random forest and support vector machines. Moreover, robustness of machine learning methods to different levels of artificially introduced noise was assessed. The open-source Caffe deep-learning framework and modern NVidia GPU units were utilized to carry out this study, allowing large number of DNN configurations to be explored.ResultsWe show that feed-forward deep neural networks are capable of achieving strong classification performance and outperform shallow methods across diverse activity classes when optimized. Hyper-parameters that were found to play critical role are the activation function, dropout regularization, number hidden layers and number of neurons. When compared to the rest methods, tuned DNNs were found to statistically outperform, with p value <0.01 based on Wilcoxon statistical test. DNN achieved on average MCC units of 0.149 higher than NB, 0.092 than kNN, 0.052 than SVM with linear kernel, 0.021 than RF and finally 0.009 higher than SVM with radial basis function kernel. When exploring robustness to noise, non-linear methods were found to perform well when dealing with low levels of noise, lower than or equal to 20%, however when dealing with higher levels of noise, higher than 30%, the Naïve Bayes method was found to perform well and even outperform at the highest level of noise 50% more sophisticated methods across several datasets.

Prediction of the Inhibitory Concentrations of Chloroquine Derivatives Using Deep Neural Networks Models.

Back-Propagation Network Improved by Conjugate Gradient Based on Genetic Algorithm in Qsar Study on Endocrine Disrupting Chemicals

Development of Predictive QSAR Models for the Substrates/inhibitors of OATP1B1 by Deep Neural Networks

QSAR Modeling for the Antimalarial Activity of 1,4-Naphthoquinonyl Derivatives As Potential Antimalarial Agents

Prediction of the Aquatic Toxicity of Aromatic Compounds to Tetrahymena Pyriformis Through Support Vector Regression

Neural networks in QSAR studies: Estimation and prediction of biological activity for N-(substituted phenoxyethyl)-cyclopropylamines

Probabilistic neural network-based on QSAR for the prediction of the diuretic activity of the active constituents of traditional Chinese medicinal herbs

Prediction of dihydrofolate reductase inhibition and selectivity using computational neural networks and linear discriminant analysis

Machine learning and deep learning approaches for enhanced prediction of hERG blockade: a comprehensive QSAR modeling study

Deep-learning: investigating deep neural networks hyper-parameters and comparison of performance to shallow methods for modeling bioactivity data

An accurate nonlinear QSAR model for the antitumor activities of chloroethylnitrosoureas using neural networks.

Accurate predictions of drugs aqueous solubility via deep learning tools

ChemNTP: Advanced Prediction of Neurotoxicity Targets for Environmental Chemicals Using a Siamese Neural Network

Comparative classification study of toxicity mechanisms using support vector machines and radial basis function neural networks

Prediction of Cytochrome P450 Inhibition Using a Deep Learning Approach and Substructure Pattern Recognition

Dual Self-Adaptive Intelligent Optimization of Feature and Hyperparameter Determination in Constructing a DNN Based QSPR Property Prediction Model

Development of Quantitative Structure-Activity Relationship Models to Predict Potential Nephrotoxic Ingredients in Traditional Chinese Medicines

High-accuracy QSAR models of narcosis toxicities of phenols based on various data partition, descriptor selection and modelling methods

A Pharmacokinetic-Pharmacodynamic Model Based on the SSA-1DCNN-Attention Network and the Semicompartment Method

Modelling enzyme inhibition toxicity of ionic liquid from molecular structure via convolutional neural network model

Novel Neural Network-Based Prediction Model for Quantifying Hydroquinone in Compost with Biosensor Measurements