The VERA software: Implementation of the acute fish toxicity endpoint and its application to pharmaceutical compounds.

The Virtual Extensive Read-Across software (VERA) is a new tool for read-across using a global similarity score, molecular groups, and structural alerts to find clusters of similar substances; these clusters are then used to identify suitable similar substances and make an assessment for the target substance. A beta version of VERA GUI is free and available at vegahub.eu; the source code of the VERA algorithm is available on GitHub. In the past we described its use to assess carcinogenicity, a classification endpoint. The aim here is to extend the automated read-across approach to assess continuous endpoints as well. We addressed acute fish toxicity. VERA evaluation on the acute fish toxicity endpoint was done on a dataset containing general substances (pesticides, industrial products, biocides, etc.), obtaining an overall R2 of 0.68. We employed the VERA algorithm also on active pharmaceutical ingredients (APIs). We included a portion of the APIs in the training dataset to predict APIs, successfully achieving an overall R2 of 0.63. VERA evaluates the assessment's reliability, and we reached an R2 of 0.78 and Root Mean Square Error (RMSE) of 0.44 for predictions with high reliability.

The VEGA Tool to Check the Applicability Domain Gives Greater Confidence in the Prediction of In Silico Models.

A sound assessment of in silico models and their applicability domain can support the use of new approach methodologies (NAMs) in chemical risk assessment and requires increasing the users' confidence in this approach. Several approaches have been proposed to evaluate the applicability domain of such models, but their prediction power still needs a thorough assessment. In this context, the VEGA tool capable of assessing the applicability domain of in silico models is examined for a range of toxicological endpoints. The VEGA tool evaluates chemical structures and other features related to the predicted endpoints and is efficient in measuring applicability domain, enabling the user to identify less accurate predictions. This is demonstrated with many models addressing different endpoints, towards toxicity of relevance to human health, ecotoxicological endpoints, environmental fate, physicochemical and toxicokinetic properties, for both regression models and classifiers.

Virtual Extensive Read-Across: A New Open-Access Software for Chemical Read-Across and Its Application to the Carcinogenicity Assessment of Botanicals.

Read-across applies the principle of similarity to identify the most similar substances to represent a given target substance in data-poor situations. However, differences between the target and the source substances exist. The present study aims to screen and assess the effect of the key components in a molecule which may escape the evaluation for read-across based only on the most similar substance(s) using a new open-access software: Virtual Extensive Read-Across (VERA). VERA provides a means to assess similarity between chemicals using structural alerts specific to the property, pre-defined molecular groups and structural similarity. The software finds the most similar compounds with a certain feature, e.g., structural alerts and molecular groups, and provides clusters of similar substances while comparing these similar substances within different clusters. Carcinogenicity is a complex endpoint with several mechanisms, requiring resource intensive experimental bioassays and a large number of animals; as such, the use of read-across as part of new approach methodologies would support carcinogenicity assessment. To test the VERA software, carcinogenicity was selected as the endpoint of interest for a range of botanicals. VERA correctly labelled 70% of the botanicals, indicating the most similar substances and the main features associated with carcinogenicity.

Development of new QSAR models for water, sediment, and soil half-life.

Checking the persistence of a chemical in the environment is extremely important. Regulations like REACH, the European one on chemicals, require the measurements or estimates of the half-life of the chemical in water, sediment, and soil. The use of non-testing methods, like quantitative structure-activity relationship (QSAR) models, is encouraged because it reduces costs and time. To our knowledge, there are very few freely available models for these properties and some are for specific chemical classes. Here, we present three new semi-quantitative models, one for each of the required environmental compartments (water, sediment, and soil). Using literature and REACH registration data, we developed three new counter-propagation artificial neural network models using the CPANNatNIC tool. We calculated the VEGA descriptors, and selected the relevant ones using an internal method in R based on the forward selection technique. The best model for each compartment was implemented in two open-source stand-alone tools, the VEGA platform, and the JANUS tool (https://www.vegahub.eu/). These models were also used by ECHA to build their PBT profiler available in the OECD QSAR toolbox (https://qsartoolbox.org/). Screening and prioritization are also our main target. The models perform well, with R2 always above 0.8 in training and validation. The only exception is the validation set of the soil compartment, with R2 0.68, that is above 0.8 only for compounds inside the applicability domain (automatically calculated by the system). The root mean square error (RMSE) is good, 0.34 or less in log units (again, for soil validation it is higher but it reaches 0.21 when considering only the compounds in the applicability domain). Compared with one of the most widely used tools, BIOWIN3, the proposed models give better results in terms of R2 and RMSE. For the classification, the performance is better for water and soil, and comparable or lower for sediment.

New Models to Predict the Acute and Chronic Toxicities of Representative Species of the Main Trophic Levels of Aquatic Environments.

To assess the impact of chemicals on an aquatic environment, toxicological data for three trophic levels are needed to address the chronic and acute toxicities. The use of non-testing methods, such as predictive computational models, was proposed to avoid or reduce the need for animal models and speed up the process when there are many substances to be tested. We developed predictive models for Raphidocelis subcapitata, Daphnia magna, and fish for acute and chronic toxicities. The random forest machine learning approach gave the best results. The models gave good statistical quality for all endpoints. These models are freely available for use as individual models in the VEGA platform and for prioritization in JANUS software.

SpheraCosmolife: a new tool for the risk assessment of cosmetic products.

A new, freely available software for cosmetic products has been designed that considers the regulatory framework for cosmetics. The software allows an overall toxicological evaluation of cosmetic ingredients without the need for additional testing and, depending on the product type, it applies defined exposure scenarios to derive risk for consumers. It takes regulatory thresholds into account and uses either experimental values, if available, or predictions. Based on the exper­imental or predicted no observed adverse effect level (NOAEL), the software can define a point of departure (POD), which is used to calculate the margin of safety (MoS) of the query chemicals. The software also provides other toxico­logical properties, such as mutagenicity, skin sensitization, and the threshold of toxicological concern (TTC) to provide an overall evaluation of the potential chemical hazard. Predictions are calculated using in silico models implemented within the VEGA software. The full list of ingredients of a cosmetic product can be processed at the same time, at the effective concentrations in the product as given by the user. SpheraCosmolife is designed as a support tool for safety assessors of cosmetic products and can be used to prioritize the cosmetic ingredients or formulations according to their potential risk to consumers. The major novelty of the tool is that it wraps a series of models (some of them new) into a single, user-friendly software system.

QSAR Models for Human Carcinogenicity: An Assessment Based on Oral and Inhalation Slope Factors.

Carcinogenicity is a crucial endpoint for the safety assessment of chemicals and products. During the last few decades, the development of quantitative structure-activity relationship ((Q)SAR) models has gained importance for regulatory use, in combination with in vitro testing or expert-based reasoning. Several classification models can now predict both human and rat carcinogenicity, but there are few models to quantitatively assess carcinogenicity in humans. To our knowledge, slope factor (SF), a parameter describing carcinogenicity potential used especially for human risk assessment of contaminated sites, has never been modeled for both inhalation and oral exposures. In this study, we developed classification and regression models for inhalation and oral SFs using data from the Risk Assessment Information System (RAIS) and different machine learning approaches. The models performed well in classification, with accuracies for the external set of 0.76 and 0.74 for oral and inhalation exposure, respectively, and r2 values of 0.57 and 0.65 in the regression models for oral and inhalation SFs in external validation. These models might therefore support regulators in (de)prioritizing substances for regulatory action and in weighing evidence in the context of chemical safety assessments. Moreover, these models are implemented on the VEGA platform and are now freely downloadable online.

Development, validation and integration of in silico models to identify androgen active chemicals.

Humans are exposed to large numbers of environmental chemicals, some of which potentially interfere with the endocrine system. The identification of potential endocrine disrupting chemicals (EDCs) has gained increasing priority in the assessment of environmental hazards. The U.S. Environmental Protection Agency (U.S. EPA) has developed the Endocrine Disruptor Screening Program (EDSP) which aims to prioritize and screen potential EDCs. The Toxicity Forecaster (ToxCast) program has generated data using in vitro high-throughput screening (HTS) assays measuring activity of chemicals at multiple points along the androgen receptor (AR) activity pathway. In the present study, using a large and diverse data set of 1667 chemicals provided by the U.S. EPA from the combined ToxCast AR assays in the framework of the Collaborative Modeling Project for Androgen Receptor Activity (CoMPARA). Two models were built using ADMET Predictor™; one is based on Artificial Neural Networks (ANNs) technology and the other uses a Support Vector Machine (SVM) algorithm; one model is a Decision Tree (DT) developed in R; and two models make use of differently combined sets of structural alerts (SAs) automatically extracted by SARpy. We used two strategies to integrate predictions from single models; one is based on a majority vote approach and the other on prediction convergence. These strategies led to enhanced statistical performance in most cases. Moreover, the majority vote approach improved prediction coverage when one or more single models were not able to provide any estimations. This study integrates multiple in silico approaches as a virtual screening tool for use in risk assessment of endocrine disrupting chemicals.

(Eco)toxicological maps: A new risk assessment method integrating traditional and in silico tools and its application in the Ledra River (Italy).

Contaminants giving rise to emerging concern like pharmaceuticals, personal care products, pesticides and Endocrine Disrupting Chemicals (EDCs) have been detected in wastewaters, as reported in the literature, but little is known about their (eco)toxicological effects and consequent human health impact. The present study aimed at overcoming this lack of information through the use of in silico methods integrated with traditional toxicological risk analysis. This is part of a pilot project involving the management of wastewater treatment plants in the Ledra River basin (Italy). We obtained data to work up a global risk assessment method combining the evaluations of health risks to humans and ecological receptors from chemical contaminants found in this specific area. The (eco)toxicological risk is expressed by a single numerical value, permitting the comparison of different sampling sites and the evaluation of future environmental and technical interventions.

Integrated strategy for mutagenicity prediction applied to food contact chemicals.

Food contamination due to unintentional leakage of chemicals from food contact materials (FCM) is a source of increasing concern. Since for many of these substances, only limited or no toxicological data are available, the development of alternative methodologies to establish rapidly and cost-efficiently level of safety concern is critical to ensure adequate consumer protection. Computational toxicology methods are considered the most promising solutions to cope with this data gap. In particular, mutagenicity assessment has a particular relevance and is a mandatory requirement for all substances released from plastic FCM, regardless how low migration and exposure are. In the present work, a strategy integrating a number of (Quantitative) Structure Activity Relationship ((Q)SAR) models for Ames mutagenicity predictions is proposed. A list of chemicals representing likely migrating moieties from FCM was selected to test the value of the newly defined strategy and the possibility to combine predictions given by the different algorithms was evaluated. In particular, a scheme to integrate mutagenicity estimations into a single final assessment was developed resulting in an increased domain of applicability. In most cases, a deeper analysis of experimental data, where available, allowed fixing misclassification errors, highlighting the importance of data curation in the development, validation and application of in silico methods. The high accuracy of the strategy provided the rationales for its application for toxicologically uncharacterized chemicals. Finally, the overall strategy of integration will be automated through its implementation into a freely available software application.

Integrating computational methods to predict mutagenicity of aromatic azo compounds.

Azo dyes have several industrial uses. However, these azo dyes and their degradation products showed mutagenicity, inducing damage in environmental and human systems. Computational methods are proposed as cheap and rapid alternatives to predict the toxicity of azo dyes. A benchmark dataset of Ames data for 354 azo dyes was employed to develop three classification strategies using knowledge-based methods and docking simulations. Results were compared and integrated with three models from the literature, developing a series of consensus strategies. The good results confirm the usefulness of in silico methods as a support for experimental methods to predict the mutagenicity of azo compounds.

A New Structure-Activity Relationship (SAR) Model for Predicting Drug-Induced Liver Injury, Based on Statistical and Expert-Based Structural Alerts.

The prompt identification of chemical molecules with potential effects on liver may help in drug discovery and in raising the levels of protection for human health. Besides in vitro approaches, computational methods in toxicology are drawing attention. We built a structure-activity relationship (SAR) model for evaluating hepatotoxicity. After compiling a data set of 950 compounds using data from the literature, we randomly split it into training (80%) and test sets (20%). We also compiled an external validation set (101 compounds) for evaluating the performance of the model. To extract structural alerts (SAs) related to hepatotoxicity and non-hepatotoxicity we used SARpy, a statistical application that automatically identifies and extracts chemical fragments related to a specific activity. We also applied the chemical grouping approach for manually identifying other SAs. We calculated accuracy, specificity, sensitivity and Matthews correlation coefficient (MCC) on the training, test and external validation sets. Considering the complexity of the endpoint, the model performed well. In the training, test and external validation sets the accuracy was respectively 81, 63, and 68%, specificity 89, 33, and 33%, sensitivity 93, 88, and 80% and MCC 0.63, 0.27, and 0.13. Since it is preferable to overestimate hepatotoxicity rather than not to recognize unsafe compounds, the model's architecture followed a conservative approach. As it was built using human data, it might be applied without any need for extrapolation from other species. This model will be freely available in the VEGA platform.

Integrating in silico models to enhance predictivity for developmental toxicity.

Application of in silico models to predict developmental toxicity has demonstrated limited success particularly when employed as a single source of information. It is acknowledged that modelling the complex outcomes related to this endpoint is a challenge; however, such models have been developed and reported in the literature. The current study explored the possibility of integrating the selected public domain models (CAESAR, SARpy and P&G model) with the selected commercial modelling suites (Multicase, Leadscope and Derek Nexus) to assess if there is an increase in overall predictive performance. The results varied according to the data sets used to assess performance which improved upon model integration relative to individual models. Moreover, because different models are based on different specific developmental toxicity effects, integration of these models increased the applicable chemical and biological spaces. It is suggested that this approach reduces uncertainty associated with in silico predictions by achieving a consensus among a battery of models. The use of tools to assess the applicability domain also improves the interpretation of the predictions. This has been verified in the case of the software VEGA, which makes freely available QSAR models with a measurement of the applicability domain.

A knowledge-based expert rule system for predicting mutagenicity (Ames test) of aromatic amines and azo compounds.

Cancer is one of the main causes of death in Western countries, and a major issue for human health. Prolonged exposure to a number of chemicals was observed to be one of the primary causes of cancer in occupationally exposed persons. Thus, the development of tools for identifying hazardous chemicals and the increase of mechanistic understanding of their toxicity is a major goal for scientific research. We constructed a new knowledge-based expert system accounting the effect of different substituents for the prediction of mutagenicity (Ames test) of aromatic amines, a class of compounds of major concern because of their widespread application in industry. The herein presented model implements a series of user-defined structural rules extracted from a database of 616 primary aromatic amines, with their Ames test outcomes, aimed at identifying mutagenic and non-mutagenic chemicals. The chemical rationale behind such rules is discussed. Besides assessing the model's ability to correctly classify aromatic amines, its predictivity was further evaluated on a second database of 354 azo dyes, another class of chemicals of major concern, whose toxicity has been predicted on the basis of the toxicity of aromatic amines potentially generated from the metabolic reduction of the azo bond. Good performance in classification on both the amine (MCC, Matthews Correlation Coefficient=0.743) and the azo dye (MCC=0.584) datasets confirmed the predictive power of the model, and its suitability for use on a wide range of chemicals. Finally, the model was compared with a series of well-known mutagenicity predicting software. The good performance of our model compared with other mutagenicity models, especially in predicting azo dyes, confirmed the usefulness of this expert system as a reliable support to in vitro mutagenicity assays for screening and prioritization purposes. The model has been fully implemented as a KNIME workflow and is freely available for downstream users.

Integrated in silico strategy for PBT assessment and prioritization under REACH.

Chemicals may persist in the environment, bioaccumulate and be toxic for humans and wildlife, posing great concern. These three properties, persistence (P), bioaccumulation (B), and toxicity (T) are the key targets of the PBT-hazard assessment. The European regulation for the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) requires assessment of PBT-properties for all chemicals that are produced or imported in Europe in amounts exceeding 10 tonnes per year, checking whether the criteria set out in REACH Annex XIII are met, so the substance should therefore be considered to have properties of very high concern. Considering how many substances can fall under the REACH regulation, there is a pressing need for new strategies to identify and screen large numbers fast and inexpensively. An efficient non-testing screening approach to identify PBT candidates is necessary, as a valuable alternative to money- and time-consuming laboratory tests and a good start for prioritization since few tools exist (e.g. the PBT profiler developed by US EPA). The aim of this work was to offer a conceptual scheme for identifying and prioritizing chemicals for further assessment and if appropriate further testing, based on their PBT-potential, using a non-testing screening approach. We integrated in silico models (using existing and developing new ones) in a final algorithm for screening and ranking PBT-potential, which uses experimental and predicted values as well as associated uncertainties. The Multi-Criteria Decision-Making (MCDM) theory was used to integrate the different values. Then we compiled a new set of data containing known PBT and non-PBT substances, in order to check how well our approach clearly differentiated compounds labeled as PBT from those labeled as non-PBT. This indicated that the integrated model distinguished between PBT from non-PBT compounds.

New Quantitative Structure-Activity Relationship Models Improve Predictability of Ames Mutagenicity for Aromatic Azo Compounds.

Existing Quantitative Structure-Activity Relationship (QSAR) models have limited predictive capabilities for aromatic azo compounds. In this study, 2 new models were built to predict Ames mutagenicity of this class of compounds. The first one made use of descriptors based on simplified molecular input-line entry system (SMILES), calculated with the CORAL software. The second model was based on the k-nearest neighbors algorithm. The statistical quality of the predictions from single models was satisfactory. The performance further improved when the predictions from these models were combined. The prediction results from other QSAR models for mutagenicity were also evaluated. Most of the existing models were found to be good at finding toxic compounds but resulted in many false positive predictions. The 2 new models specific for this class of compounds avoid this problem thanks to a larger set of related compounds as training set and improved algorithms.

New clues on carcinogenicity-related substructures derived from mining two large datasets of chemical compounds.

In this study, new molecular fragments associated with genotoxic and nongenotoxic carcinogens are introduced to estimate the carcinogenic potential of compounds. Two rule-based carcinogenesis models were developed with the aid of SARpy: model R (from rodents' experimental data) and model E (from human carcinogenicity data). Structural alert extraction method of SARpy uses a completely automated and unbiased manner with statistical significance. The carcinogenicity models developed in this study are collections of carcinogenic potential fragments that were extracted from two carcinogenicity databases: the ANTARES carcinogenicity dataset with information from bioassay on rats and the combination of ISSCAN and CGX datasets, which take into accounts human-based assessment. The performance of these two models was evaluated in terms of cross-validation and external validation using a 258 compound case study dataset. Combining R and H predictions and scoring a positive or negative result when both models are concordant on a prediction, increased accuracy to 72% and specificity to 79% on the external test set. The carcinogenic fragments present in the two models were compared and analyzed from the point of view of chemical class. The results of this study show that the developed rule sets will be a useful tool to identify some new structural alerts of carcinogenicity and provide effective information on the molecular structures of carcinogenic chemicals.

A new integrated in silico strategy for the assessment and prioritization of persistence of chemicals under REACH.

The fact that chemicals can be recalcitrant and persist in the environment arouses concern since their effects may seriously harm human and environmental health.We compiled three datasets containing half-life (HL) data on sediment, soil and water compartments in order to build in silico models and, finally, an integrated strategy for predicting persistence to be used within the EU legislation Registration, Evaluation, Authorisation and restriction of CHemicals (REACH). After splitting the datasets into training (80%) and test sets (20%), we developed models for each compartment using the k-nearest neighbor algorithm (k-NN). Accuracy was higher than 0.79 and 0.76 respectively in the training and test sets for all three compartments. To support the k-NN predictions, we identified some structural alerts, using SARpy software, with a high-true positive percentage in the test set and some chemical classes related to persistence using the software IstChemFeat. All these results were combined to build an integratedmodel and to reach to an overall conclusion (based on assessment and reliability) on the persistenceof the substance. The results on the external validation set were very encouraging and support the idea that this tool can be used successfully for regulatory purposes and to prioritize substances.

Predicting persistence in the sediment compartment with a new automatic software based on the k-Nearest Neighbor (k-NN) algorithm.

The ability of a substance to resist degradation and persist in the environment needs to be readily identified in order to protect the environment and human health. Many regulations require the assessment of persistence for substances commonly manufactured and marketed. Besides laboratory-based testing methods, in silico tools may be used to obtain a computational prediction of persistence. We present a new program to develop k-Nearest Neighbor (k-NN) models. The k-NN algorithm is a similarity-based approach that predicts the property of a substance in relation to the experimental data for its most similar compounds. We employed this software to identify persistence in the sediment compartment. Data on half-life (HL) in sediment were obtained from different sources and, after careful data pruning the final dataset, containing 297 organic compounds, was divided into four experimental classes. We developed several models giving satisfactory performances, considering that both the training and test set accuracy ranged between 0.90 and 0.96. We finally selected one model which will be made available in the near future in the freely available software platform VEGA. This model offers a valuable in silico tool that may be really useful for fast and inexpensive screening.

Hierarchical Rules for Read-Across and In Silico Models of Mutagenicity.

A broad set of rules has been implemented within the ToxRead software for read-across of chemicals for bacterial mutagenicity. These rules were obtained by manually analyzing more than 6000 chemicals and the associated chemical classes. A hierarchy of rules was established to identify those most specifically relating to the target compounds, linked in sequence to the other, more generic ones, which may match with the target compound. Rules related to both mutagenicity and lack of mutagenicity were found. Some of the latter are exceptions to the mutagenicity rules, while others are modulators of activity. These rules can also be used to predict mutagenicity, offering good performance.