Prediction Of Chemical Toxicity Research Articles

An in silico algal toxicity model with a wide applicability potential for industrial chemicals and pharmaceuticals.

The authors modeled the 72-h algal toxicity data of hundreds of chemicals with different modes of action as a function of chemical structures. They developed mode of action-based local quantitative structure-toxicity relationship (QSTR) models for nonpolar and polar narcotics as well as a global QSTR model with a wide applicability potential for industrial chemicals and pharmaceuticals. The present study rigorously evaluated the generated models, meeting the Organisation for Economic Co-operation and Development principles of robustness, validity, and transparency. The proposed global model had a broad structural coverage for the toxicity prediction of diverse chemicals (some of which are high-production volume chemicals) with no experimental toxicity data. The global model is potentially useful for endpoint predictions, the evaluation of algal toxicity screening, and the prioritization of chemicals, as well as for the decision of further testing and the development of risk-management measures in a scientific and regulatory frame. Environ Toxicol Chem 2017;36:1012-1019. © 2016 SETAC.

Prediction of developmental chemical toxicity based on gene networks of human embryonic stem cells.

Predictive toxicology using stem cells or their derived tissues has gained increasing importance in biomedical and pharmaceutical research. Here, we show that toxicity category prediction by support vector machines (SVMs), which uses qRT-PCR data from 20 categorized chemicals based on a human embryonic stem cell (hESC) system, is improved by the adoption of gene networks, in which network edge weights are added as feature vectors when noisy qRT-PCR data fail to make accurate predictions. The accuracies of our system were 97.5–100% for three toxicity categories: neurotoxins (NTs), genotoxic carcinogens (GCs) and non-genotoxic carcinogens (NGCs). For two uncategorized chemicals, bisphenol-A and permethrin, our system yielded reasonable results: bisphenol-A was categorized as an NGC, and permethrin was categorized as an NT; both predictions were supported by recently published papers. Our study has two important features: (i) as the first study to employ gene networks without using conventional quantitative structure-activity relationships (QSARs) as input data for SVMs to analyze toxicogenomics data in an hESC validation system, it uses additional information of gene-to-gene interactions to significantly increase prediction accuracies for noisy gene expression data; and (ii) using only undifferentiated hESCs, our study has considerable potential to predict late-onset chemical toxicities, including abnormalities that occur during embryonic development.

Analysis of the chemical toxicity effects using the enrichment of Gene Ontology terms and KEGG pathways

BackgroundChemical toxicity is one of the major barriers for designing and detecting new chemical entities during drug discovery. Unexpected toxicity of an approved drug may lead to withdrawal from the market and significant loss of the associated costs. Better understanding of the mechanisms underlying various toxicity effects can help eliminate unqualified candidate drugs in early stages, allowing researchers to focus their attention on other more viable candidates. MethodsIn this study, we aimed to understand the mechanisms underlying several toxicity effects using Gene Ontology (GO) terms and KEGG pathways. GO term and KEGG pathway enrichment theories were adopted to encode each chemical, and the minimum redundancy maximum relevance (mRMR) was used to analyze the GO terms and the KEGG pathways. Based on the feature list obtained by the mRMR method, the most related GO terms and KEGG pathways were extracted. ResultsSome important GO terms and KEGG pathways were uncovered, which were concluded to be significant for determining chemical toxicity effects. ConclusionsSeveral GO terms and KEGG pathways are highly related to all investigated toxicity effects, while some are specific to a certain toxicity effect. General significanceThe findings in this study have the potential to further our understanding of different chemical toxicity mechanisms and to assist scientists in developing new chemical toxicity prediction algorithms. This article is part of a Special Issue entitled “System Genetics” Guest Editor: Dr. Yudong Cai and Dr. Tao Huang.

Inside Cover Image, Volume 6, Issue 2

The inside cover image, by Vladimir B. Bajic and Arwa B. Raies, is based on the Advanced Review In silico toxicology: computational methods for prediction of chemical toxicity, DOI: 10.1002/wcms.1240. image

Evidence-based absorption, distribution, metabolism, excretion (ADME) and its interplay with alternative toxicity methods.

ADME (absorption, distribution, metabolism, elimination) has rapidly evolved over the past two decades, creating a unique interdisciplinary interface between medicinal chemists, biologists, formulators, toxicologists, clinicians, and regulators across industries, but has advanced most rapidly in the pharmaceutical industry. The implementation of ADME profiling of drug candidates, in conjunction with biological efficacy and safety optimization, has dramatically reduced pharmacokinetic drug failures in clinical trials and has become a lingua franca between disciplines that are involved in drug development. This article briefly reviews the basics and current state-of-the-art of ADME and the major lessons from the pharmaceutical industry on its efficient use, points out the importance of defining ADME properties leading to toxicity across industries for safety and toxicity prediction of chemicals, and raises the issues of quality, reliability, and reproducibility of tests and inclusion of ADME under the umbrella of evidence-based toxicology. Increasingly, in vitro results are used to inform ADME assessments and computer modeling. The aspects of kinetics of substances in cellular models themselves, however, are still too often neglected. ADME information will play a critical role in establishing quantitative in vitro to in vivo extrapolations (QIVIVE), integrated testing strategies, and systems toxicology approaches.

In silico toxicity prediction of chemicals from EPA toxicity database by kernel fusion-based support vector machines

There is a great need to assess the harmful effects or toxicities of chemicals to which man is exposed. In the present paper, the kernel fusion technique, together with the state-of-the-art support vector machine (SVM) algorithm, was developed to classify the toxicity of chemicals from Distributed Structure-Searchable Toxicity (DSSTox) database network. In this method, different kernels were firstly constructed by applying different molecular fingerprint systems, including FP2, FP4 and MACCS, and then these kernels were integrated to form a new fused kernel strictly under the algorithmic framework of kernel methods. The fused kernel can accurately measure the similarities of molecules for the toxicity classification, taking advantage of the complementarity in multiple kernels and therefore improving the prediction performance. Two model validation approaches, five-fold cross-validation and independent validation set, were used for assessing the predictive capability of our developed models. The obtained results indicate that the kernel fusion-based SVM gave the best prediction ability compared to single fingerprint kernels, and therefore could be regarded as a very promising and alternative modeling approach for potential toxicity prediction of chemicals.

Quantitative Regression Models for the Prediction of Chemical Properties by an Efficient Workflow.

Rapid safety assessment is more and more needed for the increasing chemicals both in chemical industries and regulators around the world. The traditional experimental methods couldn't meet the current demand any more. With the development of the information technology and the growth of experimental data, in silico modeling has become a practical and rapid alternative for the assessment of chemical properties, especially for the toxicity prediction of organic chemicals. In this study, a quantitative regression workflow was built by KNIME to predict chemical properties. With this regression workflow, quantitative values of chemical properties can be obtained, which is different from the binary-classification model or multi-classification models that can only give qualitative results. To illustrate the usage of the workflow, two predictive models were constructed based on datasets of Tetrahymena pyriformis toxicity and Aqueous solubility. The qcv (2) and qtest (2) of 5-fold cross validation and external validation for both types of models were greater than 0.7, which implies that our models are robust and reliable, and the workflow is very convenient and efficient in prediction of various chemical properties.

Bridging from syntactic to statistical methods: Classification with automatically segmented features from sequences

To integrate the benefits of statistical methods into syntactic pattern recognition, a Bridging Approach is proposed: (i) acquisition of a grammar per recognition class; (ii) comparison of the obtained grammars in order to find substructures of interest represented as sequences of terminal and/or non-terminal symbols and filling the feature vector with their counts; (iii) hierarchical feature selection and hierarchical classification, deducing and accounting for the domain taxonomy. The bridging approach has the benefits of syntactic methods: preserves structural relations and gives insights into the problem. Yet, it does not imply distance calculations and, thus, saves a non-trivial task-dependent design step. Instead it relies on statistical classification from many features. Our experiments concern a difficult problem of chemical toxicity prediction. The code and the data set are open-source.

Establishing Good Computer Modelling Practice (GCMP) in the Prediction of Chemical Toxicity.

This paper suggests guidelines for good computer modelling practice (GCMP) when predicting chemical toxicity, with similar purposes to those for Good Laboratory Practice (GLP). The purpose of GCMP is not to specify what should be delivered with models or predictions but to set out what must be done to ensure that work can be audited, on site, in a way analogous to the auditing of studies conforming to GLP; it is intended to confirm that work has been done properly, as distinct from providing advice on how to do it. Comments are made on the guidelines and how they might be followed, based on practical experience with the implementation of such a scheme in the development of knowledge-based and quantitative structure activity relationship models. It is hoped that publication of this paper will encourage wider discussion of the subject leading to adoption of measures to ensure the trustworthiness of computer modelling work that is carried out in connection with regulatory submissions.

In silico prediction of chemical aquatic toxicity with chemical category approaches and substructural alerts

Aquatic toxicity is an important endpoint in the evaluation of chemically adverse effects on ecosystems.

In silico prediction of chemical toxicity on avian species using chemical category approaches

Avian species are sensitive to pesticides and industrial chemicals, and hence used as model species in evaluation of chemical toxicity. In present study, we assessed the toxicity of more than 663 diverse chemicals on 17 avian species. All the chemicals were classified into three categories, i.e. highly toxic, slightly toxic and non-toxic, based on the toxicity classification criteria of the United States Environmental Protection Agency (EPA). To evaluate these chemicals, the toxicity prediction models were built using chemical category approaches with molecular descriptors and five commonly used fingerprints, in which five machine learning methods were performed on two standard test species: aquatic bird mallard duck and terrestrial bird northern bobwhite quail. The support vector machine (SVM) method with Pubchem fingerprint performed best as revealed by 5-fold cross-validation and the external validation set on Japanese quail. No species difference existed in our database despite several chemicals with different toxicity on some avian species. The best model had an overall accuracy at 0.851 for the prediction of toxicity on avian species, which outperformed the work of Mazzatorta et al. Furthermore, several representative substructures for characterizing avian toxicity were identified via information gain (IG) method. This study would provide a new tool for chemical safety assessment.

SEURAT-1 liver gold reference compounds: a mechanism-based review.

There is an urgent need for the development of alternative methods to replace animal testing for the prediction of repeat dose chemical toxicity. To address this need, the European Commission and Cosmetics Europe have jointly funded a research program for 'Safety Evaluation Ultimately Replacing Animal Testing.' The goal of this program was the development of in vitro cellular systems and associated computational capabilities for the prediction of hepatic, cardiac, renal, neuronal, muscle, and skin toxicities. An essential component of this effort is the choice of appropriate reference compounds that can be used in the development and validation of assays. In this review, we focus on the selection of reference compounds for liver pathologies in the broad categories of cytotoxicity and lipid disorders. Mitochondrial impairment, oxidative stress, and apoptosis are considered under the category of cytotoxicity, while steatosis, cholestasis, and phospholipidosis are considered under the category of lipid dysregulation. We focused on four compound classes capable of initiating such events, i.e., chemically reactive compounds, compounds with specific cellular targets, compounds that modulate lipid regulatory networks, and compounds that disrupt the plasma membrane. We describe the molecular mechanisms of these compounds and the cellular response networks which they elicit. This information will be helpful to both improve our understanding of mode of action and help in the selection of appropriate mechanistic biomarkers, allowing us to progress the development of animal-free models with improved predictivity to the human situation.

Development of in silico models for predicting LSER molecular parameters and for acute toxicity prediction to fathead minnow (Pimephales promelas)

Many chemicals with toxic effects to aquatic species are produced every year. To date, linear solvation energy relationship (LSER) models for toxicity prediction to aquatic species are limited to non-polar and polar narcotic compounds. In this study, the Verhaar scheme was used to classify chemicals into five modes of toxic actions. LSER models for predicting acute toxicity to fathead minnow were developed by identifying chemical functional groups that influence toxicity prediction of reactive chemicals. Moreover, the predictive models that can be used to estimate LSER molecular parameters have been developed by using quantum chemical and Dragon descriptors. All the predictive models were developed following the OECD guidelines for QSAR model development and validation, with a satisfactory goodness-of-fit, robustness and predictive ability. The McGowans volume was the most significant descriptor in the toxicity models. This study also inferred that, compounds with carbonyl group have different behaviors such that some can biodegrade in the organism while others do not biodegrade, which might be the reason for the difficulties in modeling the acute toxicity of reactive chemicals.

Quantitative High Content Imaging of Cellular Adaptive Stress Response Pathways in Toxicity for Chemical Safety Assessment

Over the past decade, major leaps forward have been made on the mechanistic understanding and identification of adaptive stress response landscapes underlying toxic insult using transcriptomics approaches. However, for predictive purposes of adverse outcome several major limitations in these approaches exist. First, the limited number of samples that can be analyzed reduces the in depth analysis of concentration-time course relationships for toxic stress responses. Second these transcriptomics analysis have been based on the whole cell population, thereby inevitably preventing single cell analysis. Third, transcriptomics is based on the transcript level, totally ignoring (post)translational regulation. We believe these limitations are circumvented with the application of high content analysis of relevant toxicant-induced adaptive stress signaling pathways using bacterial artificial chromosome (BAC) green fluorescent protein (GFP) reporter cell-based assays. The goal is to establish a platform that incorporates all adaptive stress pathways that are relevant for toxicity, with a focus on drug-induced liver injury. In addition, cellular stress responses typically follow cell perturbations at the subcellular organelle level. Therefore, we complement our reporter line panel with reporters for specific organelle morphometry and function. Here, we review the approaches of high content imaging of cellular adaptive stress responses to chemicals and the application in the mechanistic understanding and prediction of chemical toxicity at a systems toxicology level.

Predictive modeling of chemical toxicity towards Pseudokirchneriella subcapitata using regression and classification based approaches

Biodiversity nurturing may be a valuable pathway in controlling chemical stress on the ecosystem. In the present work, in silico studies have been performed to develop regression based quantitative structure toxicity relationship (QSTR) models using a data set containing 105 organic chemicals for the prediction of 48-h chemical toxicity towards Pseudokirchneriella subcapitata. Classification based linear discriminant analysis (LDA) was also performed to distinguish chemicals into toxic and nontoxic groups using the same data set. The developed models were found to possess good predictive quality in terms of internal, external and overall validation parameters. The regression based QSTR model suggests that second order molecular connectivity index (molecular size and lipophilicity), density (aromaticity), relative shape of molecules (cyclicity/aromaticity), and specific molecular fragments of the chemicals are important properties of chemicals to exert their toxicity on P. subcapitata. The classification based LDA QSTR model suggested that fused ring aromatic systems, secondary carbon atom fragments, second order valence molecular connectivity indices (molecular size and branching) and molecular weight are the distinguishing features to differentiate chemicals into toxic and nontoxic groups.

Toxicity of phenmedipham and carbendazim to Enchytraeus crypticus and Eisenia andrei (Oligochaeta) in Mediterranean soils

PurposeThe main objective of the present study was to evaluate the toxicity of two reference chemicals, Carbendazim and Phenmedipham, for the compostworm Eisenia andrei (effects of Carbendazim) and the potworm Enchytraeus crypticus (effects of Phenmedipham) in 12 Mediterranean soils with contrasting soil properties. The observed toxicity was also compared to that obtained for OECD standard soil, used as a control.Materials and methodsThe soils were selected to be representative for the Mediterranean region and to cover a broad range of soil properties. The evaluated endpoints were avoidance behavior and reproduction. Soils were also assembled in two groups according to their pedological properties.Results and discussionToxicity benchmarks (AC50s) obtained for E. andrei avoidance behavior in carbendazim-contaminated soils were generally higher for sandy soils with low pH. The toxic effects on the reproduction of the compostworms were similar in the six tested soils, indicating a low influence of soil properties. The avoidance response of E. crypticus towards Phenmedipham was generally highly variable in all tested soils. Even though, a higher toxicity was observed for more acidic soils. The EC50s for reproduction of the latter species varied by a factor of 9 and Phenmedipham toxicity also tended to be increasing in soils with lower pH, except for the soils with extreme organic matter content (0.6 and 5.8%).ConclusionsA soil effect on chemical toxicity was clearly confirmed, highlighting the influence that test soils can have in site-specific ecological risk assessment. Despite some relationships between soil properties and toxicity were outlined, a clear and statistically significant prediction of chemical toxicity could not be established. The range of soil properties was probably narrow to give clearer and more consistent insights on their influence. For the four groups of tests, the toxicity observed for OECD soil was either similar, lower, or generally higher if compared with Mediterranean soils. Moreover, it did represent neither the organic matter content found in Mediterranean soils nor their textural classes.

A CTD-Pfizer collaboration: manual curation of 88 000 scientific articles text mined for drug-disease and drug-phenotype interactions

Improving the prediction of chemical toxicity is a goal common to both environmental health research and pharmaceutical drug development. To improve safety detection assays, it is critical to have a reference set of molecules with well-defined toxicity annotations for training and validation purposes. Here, we describe a collaboration between safety researchers at Pfizer and the research team at the Comparative Toxicogenomics Database (CTD) to text mine and manually review a collection of 88 629 articles relating over 1 200 pharmaceutical drugs to their potential involvement in cardiovascular, neurological, renal and hepatic toxicity. In 1 year, CTD biocurators curated 2 54 173 toxicogenomic interactions (1 52 173 chemical–disease, 58 572 chemical–gene, 5 345 gene–disease and 38 083 phenotype interactions). All chemical–gene–disease interactions are fully integrated with public CTD, and phenotype interactions can be downloaded. We describe Pfizer’s text-mining process to collate the articles, and CTD’s curation strategy, performance metrics, enhanced data content and new module to curate phenotype information. As well, we show how data integration can connect phenotypes to diseases. This curation can be leveraged for information about toxic endpoints important to drug safety and help develop testable hypotheses for drug–disease events. The availability of these detailed, contextualized, high-quality annotations curated from seven decades’ worth of the scientific literature should help facilitate new mechanistic screening assays for pharmaceutical compound survival. This unique partnership demonstrates the importance of resource sharing and collaboration between public and private entities and underscores the complementary needs of the environmental health science and pharmaceutical communities.Database URL: http://ctdbase.org/

Some case studies on application of “rm2” metrics for judging quality of quantitative structure–activity relationship predictions: Emphasis on scaling of response data

Quantitative structure-activity relationship (QSAR) techniques have found wide application in the fields of drug design, property modeling, and toxicity prediction of untested chemicals. A rigorous validation of the developed models plays the key role for their successful application in prediction for new compounds. The r(m)(2) metrics introduced by Roy et al. have been extensively used by different research groups for validation of regression-based QSAR models. This concept has been further advanced here with introduction of scaling of response data prior to computation of r(m)(2). Further, a web application (accessible from http://aptsoftware.co.in/rmsquare/ and http://203.200.173.43:8080/rmsquare/) for calculation of the r(m)(2) metrics has been introduced here. The present study reports that the web application can be easily used for computation of r(m)(2) metrics provided observed and QSAR-predicted data for a set of compounds are available. Further, scaling of response data is recommended prior to r(m)(2) calculation.

Application of the Concentration Addition Model in the Assessment of Chemical Mixture Toxicity

Various chemical pollutants are always existing as mixtures in real environment. Currently, the assessment and prediction of chemical mixture toxicity is the hotspots and difficulties in environmental chemistry. To accurately assess and predict the toxicity of a chemical mixture, it is necessary to validate whether the toxicities of various components in the mix- ture is additive or not. Three common additive reference models, the effect summation (ES), concentration addition (CA), and independent action (IA), are available to determine the toxicity interaction. Synergism or antagonism between the compo- nents in a mixture can be identified if the observed toxicity of the mixture deviates from the prediction in terms of ES, CA, or IA. The resulting interaction type (synergism or antagonism) may be inconsistent according to those reference models ap- plied. Although the ES model is the earliest application model proposed to assess and predict mixture toxicity, its application in environmental chemistry was affected due to its limit in the interpretation of the so-called sham combination constructed by the same compound. The IA model is suitable to model the toxicities of mixtures consisting of the components showing dissimilar modes of actions. The CA can predict the toxicities of the mixtures consisting of the chemicals with similar modes of actions and can rationally interpret the sham combination which is impossible to be depicted by the ES model. The CA is therefore often considered as a standard additive model for the toxicity prediction of a chemical mixture. However, the CA is only a pragmatic model because it has no solid theory basis and no direct connection with the mechanism of action resulting in toxicity. Furthermore, there are so-called predictive blind zones in some concentration intervals on the concentration- response curve where the toxicities of mixtures cannot be predicted by the CA. So, it is necessary to carefully use the CA model.

In silico toxicity prediction by support vector machine and SMILES representation-based string kernel

There is a great need to assess the harmful effects or toxicities of chemicals to which man is exposed. In the present paper, the simplified molecular input line entry specification (SMILES) representation-based string kernel, together with the state-of-the-art support vector machine (SVM) algorithm, were used to classify the toxicity of chemicals from the US Environmental Protection Agency Distributed Structure-Searchable Toxicity (DSSTox) database network. In this method, the molecular structure can be directly encoded by a series of SMILES substrings that represent the presence of some chemical elements and different kinds of chemical bonds (double, triple and stereochemistry) in the molecules. Thus, SMILES string kernel can accurately and directly measure the similarities of molecules by a series of local information hidden in the molecules. Two model validation approaches, five-fold cross-validation and independent validation set, were used for assessing the predictive capability of our developed models. The results obtained indicate that SVM based on the SMILES string kernel can be regarded as a very promising and alternative modelling approach for potential toxicity prediction of chemicals.