Human health and animal health risk assessment of combined exposure to multiple chemicals use the same steps as single-substance risk assessment, namely problem formulation, exposure assessment, hazard assessment and risk characterisation. The main unique feature of combined RA is the assessment of combined exposure, toxicity and risk. Recently, the Scientific Committee of the European Food Safety Authority (EFSA) published two relevant guidance documents. The first one “Harmonised methodologies for the human health, animal health and ecological risk assessment of combined exposure to multiple chemicals” provides principles and explores methodologies for all steps of risk assessment together with a reporting table. This guidance supports also the default assumption that dose addition is applied for combined toxicity of the chemicals unless evidence for response addition or interactions (antagonism or synergism) is available. The second guidance document provides an account of the scientific criteria to group chemicals in assessment groups using hazard-driven criteria and prioritisation methods, i.e., exposure-driven and risk-based approaches. This manuscript describes such principles, provides a brief description of EFSA’s guidance documents, examples of applications in the human health and animal health area and concludes with a discussion on future challenges in this field.

Dimensionality reduction techniques are crucial for enabling deep learning driven quantitative structure-activity relationship (QSAR) models to navigate higher dimensional toxicological spaces, however the use of specific techniques is often arbitrary and poorly explored. Six dimensionality techniques (both linear and non-linear) were hence applied to a higher dimensionality mutagenicity dataset and compared in their ability to power a simple deep learning driven QSAR model, following grid searches for optimal hyperparameter values. It was found that comparatively simpler linear techniques, such as principal component analysis (PCA), were sufficient for enabling optimal QSAR model performances, which indicated that the original dataset was at least approximately linearly separable (in accordance with Cover’s theorem). However certain non-linear techniques such as kernel PCA and autoencoders performed at closely comparable levels, while (especially in the case of autoencoders) being more widely applicable to potentially non-linearly separable datasets. Analysis of the chemical space, in terms of XLogP and molecular weight, uncovered that the vast majority of testing data occurred within the defined applicability domain, as well as that certain regions were measurably more problematic and antagonised performances. It was however indicated that certain dimensionality reduction techniques were able to facilitate uniquely beneficial navigations of the chemical space.

Novel Quantitative Structure-Activity Relationship (QSAR) models were constructed using Graph Convolutional Networks (GCNs), to predict Drug-Induced Liver Injury (DILI), Drug-Induced Renal Injury (DIRI) and Drug-Induced Cardiotoxicity (DICT) of Selective Androgen Receptor Modulators (SARMs) – an emerging class of performance-enhancing drugs. Prior to training on DILI, DIRI and DICT datasets, the GCN QSAR models were first pre-trained on a variety of unrelated biomedical assay datasets, as an attempt to improve model performance via transfer learning. The success of the transfer learning was mixed; model performances were measurably improved via pre-training on certain datasets, by statistically weak increases. The optimal final QSAR models achieved overall accuracy scores of 68% for DILI (no significant improvement via ensemble modelling), 76% for DIRI (improved to 77% via ensemble modelling) and 65% for DICT (improved to 67% via ensemble modelling). Application of the most optimal singular models to a dataset of 25 SARMs predicted that 21 of the 25 SARMs are either DILI-positive, DIRI-positive, or both – which raises concern, given the rising use of SARMs. All SARMs except for one were predicted as DICT-negative. A novel definition of the Applicability Domain (AD) was used, intended for close relevance to the models, via generating three-dimensional graph embeddings, for each model. Convex hulls were fitted around training data embeddings, with a ±10% buffer, defining the AD as the region of embedded chemical space covered by the convex hull, for a given model. Subsequent analysis found that a majority of DILI, DIRI and DICT testing data lay within the AD, alongside a majority of the SARMs – adding consensus to the reliability of the predictions.