Evaluation of SARs for the prediction of skin irritation/corrosion potential: structural inclusion rules in the BfR decision support system

AttentiveSkin: To Predict Skin Corrosion/Irritation Potentials of Chemicals via Explainable Machine Learning Methods

Skin Corrosion/Irritation (Corr./Irrit.) has long been a health hazard in the Globally Harmonized System (GHS). Several in silico models have been built to predict Skin Corr./Irrit. as an alternative to the increasingly restricted animal testing. However, current studies are limited by data amount/quality and model availability. To address these issues, we compiled a traceable consensus GHS data set comprising 731 Corr., 1283 Irrit., and 1205 negative (Neg.) samples from 6 governmental databases and 2 external data sets. Then, a series of binary classifiers were developed with five machine learning (ML) algorithms and six molecular representations. For 10-fold cross-validation, the best Corr. vs Neg. classifier achieved an Area Under the Receiver Operating Characteristic Curve (AUC) of 97.1%, while the best Irrit. vs Neg. classifier achieved an AUC of 84.7%. Compared with existing in silico tools on external validation, our Attentive FP classifiers showed the highest metrics on Corr. vs Neg. and the second highest accuracy on Irrit. vs Neg. The SHapley Additive exPlanation approach was further applied to figure out important molecular features, and the attention weights were visualized to perform interpretable prediction. Structural alerts associated with Skin Corr./Irrit. were also identified. The interpretable Attentive FP classifiers were integrated into the software AttentiveSkin at https://github.com/BeeBeeWong/AttentiveSkin. The conventional ML classifiers are also provided on our platform admetSAR at http://lmmd.ecust.edu.cn/admetsar2/. Considering the data deficiency and the limited model availability of Skin Corr./Irrit., we believe that our data set and models could facilitate chemical safety assessment and relevant studies.

Machine-learning based prediction models for assessing skin irritation and corrosion potential of liquid chemicals using physicochemical properties by XGBoost

Skin irritation test is an essential part of the safety assessment of chemicals. Recently, computational models to predict the skin irritation draw attention as alternatives to animal testing. We developed prediction models on skin irritation/corrosion of liquid chemicals using machine learning algorithms, with 34 physicochemical descriptors calculated from the structure. The training and test dataset of 545 liquid chemicals with reliable in vivo skin hazard classifications based on UN Globally Harmonized System [category 1 (corrosive, Cat 1), 2 (irritant, Cat 2), 3 (mild irritant, Cat 3), and no category (nonirritant, NC)] were collected from public databases. After the curation of input data through removal and correlation analysis, every model was constructed to predict skin hazard classification for liquid chemicals with 22 physicochemical descriptors. Seven machine learning algorithms [Logistic regression, Naïve Bayes, k-nearest neighbor, Support vector machine, Random Forest, Extreme gradient boosting (XGB), and Neural net] were applied to ternary and binary classification of skin hazard. XGB model demonstrated the highest accuracy (0.73-0.81), sensitivity (0.71-0.92), and positive predictive value (0.65-0.81). The contribution of physicochemical descriptors to the classification was analyzed using Shapley Additive exPlanations plot to provide an insight into the skin irritation of chemicals.

Supplementary Information

The online version contains supplementary material available at 10.1007/s43188-022-00168-8.

Computational prediction of photosensitizers’ toxicity

The percentage of failures in late pharmaceutical development due to toxicity has increased dramatically over the last decade or so, resulting in increased demand for new methods to rapidly and reliably predict the toxicity of compounds. Today, computational toxicology can be used in every phase of drug discovery and development, from profiling large libraries early on, to predicting off-target effects in the mid-discovery phase, and to assess potential mutagenic impurities in development and degradants as part of life-cycle management. In this study, for the first time, in silico approaches were used to analyze the possible dark toxicity of photosensitive systems based on chlorin e6 and assessed possible toxicity of these compositions. By applying quantitative structure-activity relationship models (QSARs) and modeling adverse outcome pathways (AOPs), a potential toxic effect of water-soluble (chlorin e6 and chlorin e6 aminoamid) and hydrophobic (tetraphenylporphyrin) photosensitizers (PS) was predicted. Particularly, PSs’ protein binding ability, reactivity to form peptide adducts, glutathione conjugation, activity in dendritic cells, and gene expression activity in keratinocytes were explored. Using a metabolism simulator, possible PS metabolites were predicted and their potential toxicity was assessed as well. It was shown that all tested porphyrin PS and their predicted metabolites possess low activity in the mentioned processes and therefore are unable to cause significant adverse toxic effects under dark conditions.

Best Practices for QSAR Model Reporting: Physical and Chemical Properties, Ecotoxicity, Environmental Fate, Human Health, and Toxicokinetics Endpoints

BACKGROUND

Quantitative and qualitative structure–activity relationships (QSARs) have been used to understand chemical behavior for almost a century. The main source of QSAR models is the scientific literature, but the open question is how well these models are documented.

OBJECTIVES

The main aim of this study was to critically analyze the publication practices of QSARs with regard to transparency, potential reproducibility, and independent verification. The focus was on the level of technical completeness of the published QSARs.

METHODS

A total of 1,533 QSAR articles reporting 79 individual endpoints, mostly in environmental and health science, were reviewed. The QSAR parameters required for technical completeness were grouped into five categories: chemical structures, experimental endpoint values, descriptor values, mathematical representation of the model, and predicted endpoint values. The data were summarized and discussed using Circos plots.

RESULTS

Altogether, 42.5% of the reviewed articles were found to be potentially reproducible. The potential reproducibility for different endpoint groups varied; the respective rates were 39% for physical and chemical properties, 52% for ecotoxicity, 56% for environmental fate, 30% for human health, and 32% for toxicokinetics. The reproducibility of QSARs is discussed and placed in the context of the reproducibility of the experimental methods. Included are 65 references to open QSAR datasets as examples of models restored from scientific articles.

DISCUSSION

Strikingly poor documentation of QSARs was observed, which reduces the transparency, availability, and consequently, the application of research results in scientific, industrial, and regulatory areas. A list of the components needed to ensure the best practices for QSAR reporting is provided, allowing long-term use and preservation of the models. This list also allows an assessment of the reproducibility of models by interested parties such as journal editors, reviewers, regulators, evaluators, and potential users. https://doi.org/10.1289/EHP3264.

The toxicity of crude 4-methylcyclohexanemethanol (MCHM): review of experimental data and results of predictive models for its constituents and a putative metabolite

Crude 4-methylcyclohexanemethanol (MCHM) is an industrial solvent used to clean coal. Approximately 10 000 gallons of a liquid mixture containing crude MCHM were accidently released into the Elk River in West Virginia in January 2014. Because of the proximity to a water treatment facility, the contaminated water was distributed to approximately 300 000 residents. In this review, experimental data and computational predictions for the toxicity for crude MCHM, distilled MCHM, its other components and its putative metabolites are presented. Crude MCHM, its other constituents and its metabolites have low to moderate acute and subchronic oral toxicity. Crude MCHM has been shown not to be a skin sensitizer below certain doses, indicating that at plausible human exposures it does not cause an allergic response. Crude MCHM and its constituents cause slight to moderate skin and eye irritation in rodents at high concentrations. These chemicals are not mutagenic and are not predicted to be carcinogenic. Several of the constituents were predicted through modeling to be possible developmental toxicants; however, 1,4-cyclohexanedimethanol, 1,4-cyclohexanedicarboxylic acid and dimethyl 1,4-cyclohexanedicarboxylate did not demonstrate developmental toxicity in rat studies. Following the spill, the Centers for Disease Control and Prevention recommended a short-term health advisory level of 1 ppm for drinking water that it determined was unlikely to be associated with adverse health effects. Crude MCHM has an odor threshold lower than 10 ppb, indicating that it could be detected at concentrations at least 100-fold less than this risk criterion. Collectively, the findings and predictions indicate that crude MCHM poses no apparent toxicological risk to humans at 1 ppm in household water.

In silico toxicology: computational methods for the prediction of chemical toxicity

Determining the toxicity of chemicals is necessary to identify their harmful effects on humans, animals, plants, or the environment. It is also one of the main steps in drug design. Animal models have been used for a long time for toxicity testing. However, in vivo animal tests are constrained by time, ethical considerations, and financial burden. Therefore, computational methods for estimating the toxicity of chemicals are considered useful. In silico toxicology is one type of toxicity assessment that uses computational methods to analyze, simulate, visualize, or predict the toxicity of chemicals. In silico toxicology aims to complement existing toxicity tests to predict toxicity, prioritize chemicals, guide toxicity tests, and minimize late-stage failures in drugs design. There are various methods for generating models to predict toxicity endpoints. We provide a comprehensive overview, explain, and compare the strengths and weaknesses of the existing modeling methods and algorithms for toxicity prediction with a particular (but not exclusive) emphasis on computational tools that can implement these methods and refer to expert systems that deploy the prediction models. Finally, we briefly review a number of new research directions in in silico toxicology and provide recommendations for designing in silico models. WIREs Comput Mol Sci 2016, 6:147-172. doi: 10.1002/wcms.1240 For further resources related to this article, please visit the WIREs website.

Current and Future Perspectives on the Development, Evaluation, and Application of in Silico Approaches for Predicting Toxicity

Exploiting non-testing approaches to predict toxicity early in the drug discovery development cycle is a helpful component in minimizing expensive drug failures due to toxicity being identified in late development or even during clinical trials. Changes in regulations in the industrial chemicals and cosmetics sectors in recent years have prompted a significant number of advances in the development, application, and assessment of non-testing approaches, such as (Q)SARs. Many efforts have also been undertaken to establish guiding principles for performing read-across within category and analogue approaches. This review offers a perspective, as taken from these sectors, of the current status of non-testing approaches, their evolution in light of the advances in high-throughput approaches and constructs such as adverse outcome pathways, and their potential relevance for drug discovery. It also proposes a workflow for how non-testing approaches could be practically integrated within testing and assessment strategies.

QSAR and Predictors of Eye and Skin Effects

In this study, the ensemble of features and training samples was examined with a collection of support vector machines. The effects of data sampling methods, ratio of positive to negative compounds, and types of base models combiner to produce ensemble models were explored. The ensemble method was applied to produce four separate in silico models to classify the labels for eye/skin corrosion (H314), skin irritation (H315), serious eye damage (H318), and eye irritation (H319), which are defined in the "Globally Harmonized System of Classification and Labelling of Chemicals". To the best of our knowledge, the training set used in this work is one of the largest (made of publicly available data) with acceptable prediction performances. These models were distributed via PaDEL-DDPredictor (http://padel.nus.edu.sg/software/padelddpredictor) that can be downloaded freely for public use.

Methods for building QSARs

Structure-activity relationship (SAR) and quantitative structure-activity relationship (QSAR) models are increasingly used in toxicology, ecotoxicology, and pharmacology for predicting the activity of the molecules from their physicochemical properties and/or their structural characteristics. However, the design of such models has many traps for unwary practitioners. Consequently, the purpose of this chapter is to give a practical guide for the computation of SAR and QSAR models, point out problems that may be encountered, and suggest ways of solving them. Attempts are also made to see how these models can be validated and interpreted.

Quantitative risk assessment methods for cancer and noncancer effects

Human health risk assessments have evolved from the more qualitative approaches to more quantitative approaches in the past decade. This has been facilitated by the improvement in computer hardware and software capability and novel computational approaches being slowly recognized by regulatory agencies. These events have helped reduce the reliance on experimental animals as well as better utilization of published animal toxicology data in deriving quantitative toxicity indices that may be useful for risk management purposes. This chapter briefly describes some of the approaches as described in the guidance documents from several of the regulatory agencies as it pertains to hazard identification and dose-response assessment of a chemical. These approaches are contrasted with more novel computational approaches that provide a better grasp of the uncertainty often associated with chemical risk assessments.

Toxmatch--a chemical classification and activity prediction tool based on similarity measures

Chemical similarity forms the underlying basis for the development of (Quantitative) Structure-Activity Relationships ((Q)SARs), expert systems and chemical groupings. Recently a new software tool to facilitate chemical similarity calculations named Toxmatch was developed. Toxmatch encodes a number of similarity indices to help in the systematic development of chemical groupings, including endpoint specific groupings and read-across, and the comparison of model training and test sets. Two rule-based classification schemes were additionally implemented, namely: the Verhaar scheme for assigning mode of action for aquatic toxicants and the BfR rulebase for skin irritation and corrosion. In this study, a variety of different descriptor-based similarity indices were used to evaluate and compare the BfR training set with respect to its test set. The descriptors utilised in this comparison were the same as those used to derive the original BfR rules i.e. the descriptors selected were relevant for skin irritation/corrosion. The Euclidean distance index was found to be the most predictive of the indices in assessing the performance of the rules.

Toxmatch-a new software tool to aid in the development and evaluation of chemically similar groups

Chemical similarity is a widely used concept in toxicology, and is based on the hypothesis that similar compounds should have similar biological activities. This forms the underlying basis for performing read-across, forming chemical groups and developing (Quantitative) Structure-Activity Relationships ((Q)SARs). Chemical similarity is often perceived as structural similarity but in fact there are a number of other approaches that can be used to assess similarity. A systematic similarity analysis usually comprises two main steps. Firstly the chemical structures to be compared need to be characterised in terms of relevant descriptors which encode their physicochemical, topological, geometrical and/or surface properties. A second step involves a quantitative comparison of those descriptors using similarity (or dissimilarity) indices. This work outlines the use of chemical similarity principles in the formation of endpoint specific chemical groupings. Examples are provided to illustrate the development and evaluation of chemical groupings using a new software application called Toxmatch that was recently commissioned by the European Chemicals Bureau (ECB), of the European Commission's Joint Research Centre. Insights from using this software are highlighted with specific focus on the prospective application of chemical groupings under the new chemicals legislation, REACH.

An evaluation of the implementation of the Cramer classification scheme in the Toxtree software

Risk assessment for most human health effects is based on the threshold of a toxicological effect, usually derived from animal experiments. The Threshold of Toxicological Concern (TTC) is a concept that refers to the establishment of a level of exposure for all chemicals below which there would be no appreciable risk to human health. When carefully applied, the TTC concept can provide a means of waiving testing based on knowledge of exposure limits. Two main approaches exist; the first of these is a General Threshold of Toxicological Concern; the second approach is a TTC in relation to structural information and/or toxicological data of chemicals. The structural scheme most routinely used is that of Cramer and co-workers from 1978. Recently this scheme was encoded into a software program called Toxtree, specifically commissioned by the European Chemicals Bureau (ECB). Here we evaluate two published datasets using Toxtree to demonstrate its concordance and highlight potential software modifications. The results were promising with an overall good concordance between the reported classifications and those generated by Toxtree. Further evaluation of these results highlighted a number of inconsistencies which were examined in turn and rationalised as far as possible. Improvements for Toxtree were proposed where appropriate. Notable of these is a necessity to update the lists of common food components and normal body constituents as these accounted for the majority of false classifications observed. Overall Toxtree was found to be a useful tool in facilitating the systematic evaluation of compounds through the Cramer scheme.

Evaluation of SARs for the prediction of eye irritation/corrosion potential: structural inclusion rules in the BfR decision support system

The proposed REACH regulation within the European Union (EU) aims to minimise the number of laboratory animals used for human hazard and risk assessment while ensuring adequate protection of human health and the environment. One way to achieve this goal is to develop non-testing methods, such as (quantitative) structure-activity relationships ([Q]SARs), suitable for identifying toxicological hazard from chemical structure and physicochemical properties alone. A database containing data submitted within the EU New Chemicals Notification procedure was compiled by the German Bundesinstitut für Risikobewertung (BfR). On the basis of these data, the BfR built a decision support system (DSS) for the prediction of several toxicological endpoints. For the prediction of eye irritation and corrosion potential, the DSS contains 31 physicochemical exclusion rules evaluated previously by the European Chemicals Bureau (ECB), and 27 inclusion rules that define structural alerts potentially responsible for eye irritation and/or corrosion. This work summarises the results of a study carried out by the ECB to assess the performance of the BfR structural rulebase. The assessment included: (a) evaluation of the structural alerts by using the training set of 1341 substances with experimental data for eye irritation and corrosion; and (b) external validation by using an independent test set of 199 chemicals. Recommendations are made for the further development of the structural rules in order to increase the overall predictivity of the DSS.

The role of the European Chemicals Bureau in promoting the regulatory use of (Q)SAR methods

Under the proposed REACH (Registration, Evaluation and Authorisation of CHemicals) legislation, (Q)SAR models and grouping methods (chemical categories and read across approaches) are expected to play a significant role in prioritising industrial chemicals for further assessment, and for filling information gaps for the purposes of classification and labelling, risk assessment and the assessment of persistent, bioaccumulative and toxic (PBT) chemicals. The European Chemicals Bureau (ECB), which is part of the European Commission's Joint Research Centre (JRC), has a well-established role in providing independent scientific and technical advice to European policy makers. The ECB also promotes consensus and capacity building on scientific and technical matters among stakeholders in the Member State authorities and industry. To promote the availability and use of (Q)SARs and related estimation methods, the ECB is carrying out a range of activities, including applied research in computational toxicology, the assessment of (Q)SAR models and methods, the development of technical guidance documents and computational tools, and the organisation of training courses. This article provides an overview of ECB activities on computational toxicology, which are intended to promote the development, validation, acceptance and use of (Q)SARs and related estimation methods, both at the European and international levels.