In silico toxicology: From structure–activity relationships towards deep learning and adverse outcome pathways

Abstract

In silico toxicology is an emerging field. It gains increasing importance as research is aiming to decrease the use of animal experiments as suggested in the 3R principles by Russell and Burch. In silico toxicology is a means to identify hazards of compounds before synthesis, and thus in very early stages of drug development. For chemical industries, as well as regulatory agencies it can aid in gap‐filling and guide risk minimization strategies. Techniques such as structural alerts, read‐across, quantitative structure–activity relationship, machine learning, and deep learning allow to use in silico toxicology in many cases, some even when data is scarce. Especially the concept of adverse outcome pathways puts all techniques into a broader context and can elucidate predictions by mechanistic insights. This article is categorized under: Structure and Mechanism > Computational Biochemistry and Biophysics Data Science > Chemoinformatics

Machine learning in drug design: Use of artificial intelligence to explore the chemical structure–biological activity relationship

“…Artificial neural networks are the main method proposed for QSAR models (Figure 4). 15 This technique is widely used by the pharmaceutical industry in the drug discovery process. As early as at the beginning of the century, scientists noted that increasing computer power can support decision making in this area.…”

“…2 Artificial neural networks providing diagnostic, identification, and organizational potential, especially for large clinical and biological datasets, are becoming increasingly used in medical science. Drug discovery, [3][4][5][6][7][8][9][10] lead optimization 11 and synthesis, 12,13 cardiological and cardiovascular diseases, [14][15][16][17][18] medical image analysis, [19][20][21][22] diabetic diseases, 23,24 oncology research, 25,26 diagnosis, for example, alteration of oscillatory brain activity as a possible biomarker for use in Alzheimer's disease diagnosis, 27 are some of the examples of AI in service of medical science (Figure 1). Computer-aided drug design is not only an interesting concept but also a business requirement.…”

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

“…In the effort to improve in silico predictions of drug toxicity, multiple computational models have been developed [147,[149][150][151][152]. These include structural alerts and rule-based models, read-across, dose-response and timeresponse models, pharmacokinetic and pharmacodynamic models, uncertainty factors models and QSAR models.…”

Drug discovery with explainable artificial intelligence

“…Several concepts of 'artificial intelligence' (AI) have been adopted for computer-assisted drug discovery. [1][2][3][4][5][6][7] Deep learning algorithms, i.e., artificial neural networks with multiple processing layers, 8 are currently receiving particular attention, owing to their capacity to model complex nonlinear input-output relationships, and perform pattern recognition and feature extraction from low-level data representations. 8,9 Certain deep learning models have been shown to match or even exceed the performance of the existing machinelearning and quantitative structure-activity relationship (QSAR) methods for drug discovery.…”