Prediction Of Chemical Toxicity Research Articles

Interpreting Neural Network Models for Toxicity Prediction by Extracting Learned Chemical Features.

Neural network models have become a popular machine-learning technique for the toxicity prediction of chemicals. However, due to their complex structure, it is difficult to understand predictions made by these models which limits confidence. Current techniques to tackle this problem such as SHAP or integrated gradients provide insights by attributing importance to the input features of individual compounds. While these methods have produced promising results in some cases, they do not shed light on how representations of compounds are transformed in hidden layers, which constitute how neural networks learn. We present a novel technique to interpret neural networks which identifies chemical substructures in training data found to be responsible for the activation of hidden neurons. For individual test compounds, the importance of hidden neurons is determined, and the associated substructures are leveraged to explain the model prediction. Using structural alerts for mutagenicity from the Derek Nexus expert system as ground truth, we demonstrate the validity of the approach and show that model explanations are competitive with and complementary to explanations obtained from an established feature attribution method.

A problem formulation framework for the application of in silico toxicology methods in chemical risk assessment

The first step in the hazard or risk assessment of chemicals should be to formulate the problem through a systematic and iterative process aimed at identifying and defining factors critical to the assessment. However, no general agreement exists on what components an in silico toxicology problem formulation (PF) should include. The present work aims to develop a PF framework relevant to the application of in silico models for chemical toxicity prediction. We modified and applied a PF framework from the general risk assessment literature to peer reviewed papers describing PFs associated with in silico toxicology models. Important gaps between the general risk assessment literature and the analyzed PF literature associated with in silico toxicology methods were identified. While the former emphasizes the need for PFs to address higher-level conceptual questions, the latter does not. There is also little consistency in the latter regarding the PF components addressed, reinforcing the need for a PF framework that enable users of in silico toxicology models to answer the central conceptual questions aimed at defining components critical to the model application. Using the developed framework, we highlight potential areas of uncertainty manifestation in in silico toxicology PF in instances where particular components are missing or implicitly described. The framework represents the next step in standardizing in silico toxicology PF component. The framework can also be used to improve the understanding of how uncertainty is apparent in an in silico toxicology PF, thus facilitating ways to address uncertainty.

Transformers enable accurate prediction of acute and chronic chemical toxicity in aquatic organisms.

Environmental hazard assessments are reliant on toxicity data that cover multiple organism groups. Generating experimental toxicity data is, however, resource-intensive and time-consuming. Computational methods are fast and cost-efficient alternatives, but the low accuracy and narrow applicability domains have made their adaptation slow. Here, we present a AI-based model for predicting chemical toxicity. The model uses transformers to capture toxicity-specific features directly from the chemical structures and deep neural networks to predict effect concentrations. The model showed high predictive performance for all tested organism groups-algae, aquatic invertebrates and fish-and has, in comparison to commonly used QSAR methods, a larger applicability domain and a considerably lower error. When the model was trained on data with multiple effect concentrations (EC50/EC10), the performance was further improved. We conclude that deep learning and transformers have the potential to markedly advance computational prediction of chemical toxicity.

In Silico Prediction of Chemical Acute Dermal Toxicity Using Explainable Machine Learning Methods.

The research on acute dermal toxicity has consistently been a crucial component in assessing the potential risks of human exposure to active ingredients in pesticides and related plant protection products. However, it is difficult to directly identify the acute dermal toxicity of potential compounds through animal experiments alone. In our study, we separately integrated 1735 experimental data based on rabbits and 1679 experimental data based on rats to construct acute dermal toxicity prediction models using machine learning and deep learning algorithms. The best models for the two animal species achieved AUC values of 78.0 and 82.0%, respectively, on 10-fold cross-validation. Additionally, we employed SARpy to extract structural alerts, and in conjunction with Shapley additive explanation and attentive FP heatmap, we identified important features and structural fragments associated with acute dermal toxicity. This approach offers valuable insights for the detection of positive compounds. Moreover, a standalone software tool was developed to make acute dermal toxicity prediction easier. In summary, our research would provide an effective tool for acute dermal toxicity evaluation of pesticides, cosmetics, and drug safety assessment.

Comparative Toxicogenomics Database: A tool to investigate the effects of environmental exposures on the etiology of Alzheimer’s disease

AbstractBackgroundMost human diseases involve interactions between genetic and environmental factors. Although the environment is implicated in most chronic diseases, the etiology and mechanisms of action underlying these diseases remain unclear. It is estimated that more than 80,000 chemicals are currently used in commerce, challenging elucidation about chemical mechanisms‐of‐action and prioritization of environmental health research. Integration of diverse data is required to understand environment‐disease associations, mechanisms of chemical action, toxicity prediction and development of effective therapeutic interventions.MethodThe Comparative Toxicogenomics Database (CTD; http://ctdbase.org) was launched publicly to address an unmet need for a public data resource dedicated exclusively to advancing understanding about environment‐disease connections. CTD provides a combination of richly annotated data describing chemical‐gene‐phenotype‐disease interactions, exposure information, and novel analysis tools that enable user‐driven discoveries about environmental influences on human health.ResultsCurrently, CTD provides >2.5 million manually curated relationships for 16,800 chemicals; 52,000 genes/proteins; 5,500 phenotypes; and 7,200 diseases. The value of CTD has been specifically leveraged for a range of chronic and neurological diseases, including Alzheimer’s disease.ConclusionThere is growing evidence of environmental influences on the etiology of a range of neuropsychiatric syndromes. This presentation will provide an overview of CTD content and functionality and how it can be used to explore environmental influences on the etiology and underlying mechanisms of Alzheimer’s disease and other disorders.

Modeling and Insights into the Structural Characteristics of Chemical Mitochondrial Toxicity.

Mitochondria are the energy metabolism center of cells and are involved in a number of other processes, such as cell differentiation and apoptosis, signal transduction, and regulation of cell cycle and cell proliferation. It is of great significance to evaluate the mitochondrial toxicity of drugs and other chemicals. In the present study, we aimed to propose easily available artificial intelligence (AI) models for the prediction of chemical mitochondrial toxicity and investigate the structural characteristics with the analysis of molecular properties and structural alerts. The consensus model achieved good predictive results with high total accuracy at 87.21% for validation sets. The models can be accessed freely via https://ochem.eu/article/158582. Besides, several commonly used chemical properties were significantly different between chemicals with and without mitochondrial toxicity. We also detected the structural alerts (SAs) responsible for mitochondrial toxicity and integrated them into the web-server SApredictor (www.sapredictor.cn). The study may provide useful tools for in silico estimation of mitochondrial toxicity and be helpful to understand the mechanisms of mitochondrial toxicity.

Utility of EPA CompTox Chemical Dashboard for Next Generation Risk Assessment and Toxicity Prediction of Chemicals

In recent years, the paradigm of toxicity science is moving toward predictive science. New approach methodologies (NAMs), including <i>in chemico</i>, <i>in silico</i>, and <i>in vitro</i> approach, provides a new feasibility to minimize the whole animal toxicity testing. To support these efforts, the U.S. Environmental Protection Agency (EPA)’s web-based CompTox Chemical Dashboard was developed and providing relevant data through a curation of diverse substances and predictive tools. In this review, we presented a comprehensive analysis of EPA CompTox Chemical dashboard’s available data, predictive tools and recent case studies using them. We also suggested a perspective of utility of EPA CompTox Chemical Dashboard for implementation of next generation risk assessment and toxicity prediction of environmental chemicals.

In silico prediction of acute chemical toxicity of biocides in marine crustaceans using machine learning

Biocides are a heterogeneous group of chemical substances intended to control the growth or kill undesired organisms. Due to their extensive use, they enter marine ecosystems via non-point sources and may pose a threat to ecologically important non-target organisms. Consequently, industries and regulatory agencies have recognized the ecotoxicological hazard potential of biocides. However, the prediction of biocide chemical toxicity on marine crustaceans has not been previously evaluated. This study aims to provide in silico models capable of classifying structurally diverse biocidal chemicals into different toxicity categories and predict acute chemical toxicity (LC50) in marine crustaceans using a set of calculated 2D molecular descriptors. The models were built following the guidelines recommended by the OECD (Organization for Economic Cooperation and Development) and validated through stringent processes (internal and external validation). Six machine learning (ML) models were built and compared (linear regression: LR; support vector machine: SVM; random forest: RF; feed-forward backpropagation-based artificial neural network: ANN; decision trees: DT and naïve Bayes: NB) for regression and classification analysis to predict toxicities. All the models displayed encouraging results with high generalisability: the feed-forward-based backpropagation method showed the best results with determination coefficient R2 values of 0.82 and 0.94, respectively, for training set (TS) and validation set (VS). For classification-based modelling, the DT model performed the best with an accuracy (ACC) of 100 % and an area under curve (AUC) value of 1 for both TS and VS. These models showed the potential to replace animal testing for the chemical hazard assessment of untested biocides if they fall within the applicability domain of the proposed models. In general, the models are highly interpretable and robust, with good predictive performance. The models also displayed a trend indicating that toxicity is largely influenced by factors such as lipophilicity, branching, non-polar bonding and saturation of molecules.

Data repository for the publication "Transformers enable accurate prediction of acute and chronic chemical toxicity in aquatic organisms"

Data repository for the publication "Transformers enable accurate prediction of acute and chronic chemical toxicity in aquatic organisms"

Machine learning-assisted data filtering and QSAR models for prediction of chemical acute toxicity on rat and mouse

Machine learning (ML) methods provide a new opportunity to build quantitative structure-activity relationship (QSAR) models for predicting chemicals’ toxicity based on large toxicity data sets, but they are limited in insufficient model robustness due to poor data set quality for chemicals with certain structures. To address this issue and improve model robustness, we built a large data set on rat oral acute toxicity for thousands of chemicals, then used ML to filter chemicals favorable for regression models (CFRM). In comparison to chemicals not favorable for regression models (CNRM), CFRM accounted for 67% of chemicals in the original data set, and had a higher structural similarity and a smaller toxicity distribution in 2–4 log10 (mg/kg). The performance of established regression models for CFRM was greatly improved, with root-mean-square deviations (RMSE) in the range of 0.45–0.48 log10 (mg/kg). Classification models were built for CNRM using all chemicals in the original data set, and the area under receiver operating characteristic (AUROC) reached 0.75–0.76. The proposed strategy was successfully applied to a mouse oral acute data set, yielding RMSE and AUROC in the range of 0.36–0.38 log10 (mg/kg) and 0.79, respectively.

Co-model for chemical toxicity prediction based on multi-task deep learning.

The toxicity of compounds is closely related to the effectiveness and safety of drug development, and accurately predicting the toxicity of compounds is one of the most challenging tasks in medicinal chemistry and pharmacology. In this paper, we construct three types of models for single and multi-tasking based on 2D and 3D descriptors, fingerprints and molecular graphs, and then validate the models with benchmark tests on the Tox21 data challenge. We found that due to the information sharing mechanism of multi-task learning, it could address the imbalance problem of the Tox21 data sets to some extent, and the prediction performance of the multi-task was significantly improved compared with the single task in general. Given the complement of the different molecular representations and modeling algorithms, we attempted to integrate them into a robust Co-Model. Our Co-Model performs well in various evaluation metrics on the test set and also achieves significant performance improvement compared to other models in the literature, which clearly demonstrates its superior predictive power and robustness.

Small Molecule Screening Strategies from Lead Identification to Validation.

Festa, F., Labaer, J. Kinase inhibitor screening in self-assembled human protein microarrays. Journal of Visualized Experiments. 152, e59886 (2019). Stockman, B. J. et al. NMR-Based activity assays for determining compound inhibition, IC50 values, artifactual activity, and whole-cell activity of nucleoside ribohydrolases. Journal of Visualized Experiments. 148, e59928 (2019). Gao, S. et al. A high-throughput assay for the prediction of chemical toxicity by automated phenotypic profiling of Caenorhabditis elegans. Journal of Visualized Experiments. 145, e59082 (2019). Axelsson, H., Almqvist, H., Seashore-Ludlow, B. Using high content imaging to quantify target engagement in adherent cells. Journal of Visualized Experiments. 141, e58670 (2018). Chorba, J. S., Galvan, A. M., Shokat, K. M. A high-throughput luciferase assay to evaluate proteolysis of the single-turnover protease PCSK9. Journal of Visualized Experiments. 138, e58265 (2018). Sullivan, C. et al. Using zebrafish models of human influenza A virus infections to screen antiviral drugs and characterize host immune cell responses. Journal of Visualized Experiments. 119, e55235 (2017). Tiemann, K., Garri, C., Wang, J., Clarke, L., Kani, K. Assessment of resistance to tyrosine kinase inhibitors by an interrogation of signal transduction pathways by antibody arrays. Journal of Visualized Experiments. 139, e57779 (2018). Radnai, L., Stremel, R. F., Sellers, J. R., Rumbaugh, G., Miller, C. A. A semi-high-throughput adaptation of the NADH-coupled ATPase assay for screening small molecule inhibitors. Journal of Visualized Experiments. 150, e60017 (2019). Nandha Premnath, P., Craig, S., McInnes, C. Development of inhibitors of protein-protein interactions through REPLACE: Application to the design and development non-ATP competitive CDK inhibitors. Journal of Visualized Experiments. 104, e52441 (2015). Chen, E. W., Ke, C. Y., Brzostek, J., Gascoigne, N. R. J., Rybakin, V. Identification of mediators of T-cell receptor signaling via the screening of chemical inhibitor libraries. Journal of Visualized Experiments. 143, e58946 (2019). Takakusagi, Y. Biosensor-based high throughput biopanning and bioinformatics analysis strategy for the global validation of drug-protein interactions. Journal of Visualized Experiments. 166, e61873 (2020). Riching, K. M., Mahan, S. D., Urh, M., Daniels, D. L. High-Throughput cellular profiling of targeted protein degradation compounds using HiBiT CRISPR cell lines. Journal of Visualized Experiments. 165, e61787 (2020).

2,5-Bis(2,2,2-trifluoroethoxy)phenyl-tethered 1,3,4-Oxadiazoles Derivatives: Synthesis, In Silico Studies, and Biological Assessment as Potential Candidates for Anti-Cancer and Anti-Diabetic Agent.

In the present work, a series of new 1-{5-[2,5-bis(2,2,2-trifluoroethoxy)phenyl]-1,3,4-oxadiazol-3-acetyl-2-aryl-2H/methyl derivatives were synthesized through a multistep reaction sequence. The compounds were synthesized by the condensation of various aldehydes and acetophenones with the laboratory-synthesized acid hydrazide, which afforded the Schiff's bases. Cyclization of the Schiff bases yielded 1,3,4-oxadiazole derivatives. By spectral analysis, the structures of the newly synthesized compounds were elucidated, and further, their anti-cancer and anti-diabetic properties were investigated. To examine the dynamic behavior of the candidates at the binding site of the protein, molecular docking experiments on the synthesized compounds were performed, followed by a molecular dynamic simulation. ADMET (chemical absorption, distribution, metabolism, excretion, and toxicity) prediction revealed that most of the synthesized compounds follow Lipinski's rule of 5. The results were further correlated with biological studies. Using a cytotoxic assay, the newly synthesized 1,3,4-Oxadiazoles were screened for their in vitro cytotoxic efficacy against the LN229 Glioblastoma cell line. From the cytotoxic assay, the compounds 5b, 5d, and 5m were taken for colony formation assay and tunnel assay have shown significant cell apoptosis by damaging the DNA of cancer cells. The in vivo studies using a genetically modified diabetic model, Drosophila melanogaster, indicated that compounds 5d and 5f have better anti-diabetic activity among the different synthesized compounds. These compounds lowered the glucose levels significantly in the tested model.

Integrative Approaches of Bioassay and Computational Analysis for Discovering Potential Bioactive Compounds and Predictive Toxicity.

Changes in eating habits are brought about by drastic changes in lifestyle and environment, and, it has been pointed out, are strongly involved in the increase in neurological diseases and onset of cancer in younger adult ages. There is a wide variety of chemical substances in food, and there is a need to analyze the effects of complex exposures on complex mechanisms of action and to develop methods for evaluating and predicting them. The power of molecular nutrition needs to create an integrated approach to human nutrition in line with the grand social challenges of diet-related illnesses. The current article aims to explore some of these areas where integration is appropriate. Therefore, in this symposium, we will introduce the contents of four performers who are conducting cutting-edge research. 1) Chemoprevention by vitamin A and its derivatives, 2) Toxicity prediction of natural compounds from a developing database of bioactive gradients from Kampo medicine, 3) Toxicity prediction of chemicals using pluripotent stem cells. 4) Detection of bioactive compounds in "Aji" or "Umami" in food. By detecting and predicting the biological activity and toxicity of chemical substances such as nutrients in foods, it will be possible to provide better molecular information on dietary components. In addition, we will introduce next-generation health and prevention methods.

Predicting Dose-Range Chemical Toxicity using Novel Hybrid Deep Machine-Learning Method.

Humans are exposed to thousands of chemicals, including environmental chemicals. Unfortunately, little is known about their potential toxicity, as determining the toxicity remains challenging due to the substantial resources required to assess a chemical in vivo. Here, we present a novel hybrid neural network (HNN) deep learning method, called HNN-Tox, to predict chemical toxicity at different doses. To develop a hybrid HNN-Tox method, we combined two neural network frameworks, the Convolutional Neural Network (CNN) and the multilayer perceptron (MLP)-type feed-forward neural network (FFNN). Combining the CNN and FCNN in the field of environmental chemical toxicity prediction is a novel approach. We developed several binary and multiclass classification models to assess dose-range chemical toxicity that is trained based on thousands of chemicals with known toxicity. The performance of the HNN-Tox was compared with other machine-learning methods, including Random Forest (RF), Bootstrap Aggregation (Bagging), and Adaptive Boosting (AdaBoost). We also analyzed the model performance dependency on varying features, descriptors, dataset size, route of exposure, and toxic dose. The HNN-Tox model, trained on 59,373 chemicals annotated with known LD50 and routes of exposure, maintained its predictive ability with an accuracy of 84.9% and 84.1%, even after reducing the descriptor size from 318 to 51, and the area under the ROC curve (AUC) was 0.89 and 0.88, respectively. Further, we validated the HNN-Tox with several external toxic chemical datasets on a large scale. The HNN-Tox performed optimally or better than the other machine-learning methods for diverse chemicals. This study is the first to report a large-scale prediction of dose-range chemical toxicity with varying features. The HNN-Tox has broad applicability in predicting toxicity for diverse chemicals and could serve as an alternative methodology approach to animal-based toxicity assessment.

Random forest algorithm-based accurate prediction of chemical toxicity to Tetrahymena pyriformis

The random forest (RF) algorithm, together with ten Dragon descriptors, was used to develop a quantitative structure–toxicity/activity relationship (QSTR/QSAR) model for a larger data set of 1792 chemical toxicity pIGC50 towards Tetrahymena pyriformis. The optimal RF (ntree =300 and mtry =3) model yielded root mean square (rms) errors of 0.261 for the training set (1434 chemicals) and 0.348 for the test set (358 chemicals). Compared with other QSTR models reported in the literature, the optimal RF model in this paper is more accurate. The feasibility of applying the RF algorithm to predict chemical toxicity pIGC50 towards Tetrahymena pyriformis has been verified.

In silico prediction of boiling point, octanol-water partition coefficient, and retention time index of polycyclic aromatic hydrocarbons through machine learning.

Polycyclic aromatic hydrocarbons (PAHs), a special class of persistent organic pollutants (POPs) with two or more aromatic rings, have received extensive attention owing to their carcinogenic, mutagenic, and teratogenic effects. Quantitative structure-property relationship (QSPR) is powerful chemometric method to correlate structural descriptors of PAHs with their physicochemical properties. In this manuscript, a QSPR study of PAHs was performed to predict their boiling point (bp), octanol-water partition coefficient (LogKow ), and retention time index (RI). In addition to traditional molecular descriptors, structural fingerprints play an important role in the correlation of the above properties. Three regression methods, partial least squares (PLS), multiple linear regression (MLR), and genetic function approximation (GFA), were used to establish QSPR models for each property of PAHs. The correlation coefficient (R2 test ) and root mean square error (RMSE) of best model were 0.980 and 24.39% (PLS), 0.979 and 35.80% (GFA), 0.926 and 22.90% (MLR) for bp, LogKow, and RI, respectively. The model proposed here can be used to estimate physicochemical properties and inform toxicity prediction of environmental chemicals.

SApredictor: An Expert System for Screening Chemicals Against Structural Alerts.

The rapid and accurate evaluation of chemical toxicity is of great significance for estimation of chemical safety. In the past decades, a great number of excellent computational models have been developed for chemical toxicity prediction. But most machine learning models tend to be “black box”, which bring about poor interpretability. In the present study, we focused on the identification and collection of structural alerts (SAs) responsible for a series of important toxicity endpoints. Then, we carried out effective storage of these structural alerts and developed a web-server named SApredictor (www.sapredictor.cn) for screening chemicals against structural alerts. People can quickly estimate the toxicity of chemicals with SApredictor, and the specific key substructures which cause the chemical toxicity will be intuitively displayed to provide valuable information for the structural optimization by medicinal chemists.

In silico prediction of chemical aquatic toxicity by multiple machine learning and deep learning approaches.

Fish is one of the model animals used to evaluate the adverse effects of a chemical exposed to the ecosystem. However, its low throughput and relevantly high expense make it impossible to test all new chemicals in manufacture. Hence, using in silico models to prioritize compounds to be tested has been widely applied in environmental risk assessment and drug discovery. In this study, we constructed the local predictive models for four fish species, including bluegill sunfish, rainbow trout, fathead minnow, and sheepshead minnow, and the global models with all four fish data. A total of 1874 unique compounds with their labels, that is, toxic (LC50 < 10ppm) or nontoxic, were collected from ECOTOX and literature. Both conventional machine learning methods and the deep learning architecture, graph convolutional network (GCN), were used to build predictive models. The classification accuracy of the best local model for each fish species was higher than 0.83. For the global models, two strategies including consistency prediction and probability threshold were adopted to improve the predictive capability at the cost of limiting applicability domain. For 63% of compounds in domain, the accuracy was around 0.97. By comparison of the deep learning and machine learning methods, we found that the single-task GCN showed specific advantages in performance, and multitask GCN showed no advantages over the conventional machine learning methods. The data and models are available on GitHub (https://github.com/ChemPredict/ChemicalAquaticToxicity).

Critical features identification for chemical chronic toxicity based on mechanistic forecast models

Facing billions of tons of pollutants entering the ocean each year, aquatic toxicity is becoming a crucial endpoint for evaluating chemical adverse effects on ecosystems. Notably, huge amount of toxic chemicals at environmental relevant doses can cause potential adverse effects. However, chronic aquatic toxicity effects of chemicals are much scarcer, especially at population level. Rotifers are highly sensitive to toxicants even at chronic low-doses and their communities are usually considered as effective indicators for assessing the status of aquatic ecosystems. Therefore, the no observed effect concentration (NOEC) for population abundance of rotifers were selected as endpoints to develop machine learning models for the prediction of chemical aquatic chronic toxicity. In this study, forty-eight binary models were built by eight types of chemical descriptors combined with six machine learning algorithms. The best binary model was 1D & 2D molecular descriptors – random trees model (RT) with high balanced accuracy (BA) (0.83 for training and 0.83 for validation set), and Matthews correlation coefficient (MCC) (0.72 for training set and 0.67 for validation set). Moreover, the optimal model identified the primary factors (SpMAD_Dzp, AMW, MATS2v) and filtered out three high alerting substructures [c1cc(Cl)cc1, CNCO, CCOP(=S)(OCC)O] influencing the chronic aquatic toxicity. These results showed that the compounds with low molecular volume, high polarity and molecular weight could contribute to adverse effects on rotifers, facilitating the deeper understanding of chronic toxicity mechanisms. In addition, forecast models had better performances than the common models embedded into ECOSAR software. This study provided insights into structural features responsible for the toxicity of different groups of chemicals and thereby allowed for the rational design of green and safer alternatives.