Comparison of multi-task convolutional neural network (MT-CNN) and a few other methods for toxicity prediction

TL;DR

This study compares multi-task CNN with other machine learning methods for toxicity prediction using physically meaningful descriptors, demonstrating that MT-CNN outperforms existing approaches on benchmark datasets.

Contribution

The paper introduces a common set of microscopic descriptors and evaluates MT-CNN against other models, showing its superior performance in toxicity prediction.

Findings

MT-CNN outperforms other methods on benchmark datasets.

Physically meaningful descriptors improve model performance.

Deep learning models like CNN and DNN are compared with traditional methods.

Abstract

Toxicity analysis and prediction are of paramount importance to human health and environmental protection. Existing computational methods are built from a wide variety of descriptors and regressors, which makes their performance analysis difficult. For example, deep neural network (DNN), a successful approach in many occasions, acts like a black box and offers little conceptual elegance or physical understanding. The present work constructs a common set of microscopic descriptors based on established physical models for charges, surface areas and free energies to assess the performance of multi-task convolutional neural network (MT-CNN) architectures and a few other approaches, including random forest (RF) and gradient boosting decision tree (GBDT), on an equal footing. Comparison is also given to convolutional neural network (CNN) and non-convolutional deep neural network (DNN) algorithms. Four benchmark toxicity data sets (i.e., endpoints) are used to evaluate various approaches. Extensive numerical studies indicate that the present MT-CNN architecture is able to outperform the state-of-the-art methods.