Improving confidence in (Q)SAR predictions under Canada’s Chemicals Management Plan – a chemical space approach$

Abstract

One of the key challenges of Canada’s Chemicals Management Plan (CMP) is assessing chemicals with limited/no empirical hazard data for their risk to human health. In some instances, these chemicals have not been tested broadly for their toxicological potency; as such, limited information exists on their potential to induce human health effects following exposure. Although (quantitative) structure activity relationship ((Q)SAR) models are able to generate predictions to address data gaps for certain toxicological endpoints, the confidence in predictions also needs to be addressed. One way to address this issue is to apply a chemical space approach. This approach uses international toxicological databases, for example, those available in the Organisation for Economic Co-operation and Development (OECD) QSAR Toolbox. The approach,assesses a model’s ability to predict the potential hazards of chemicals that have limited hazard data that require assessment under the CMP when compared to a larger, data-rich chemical space that is structurally similar to chemicals of interest. This evaluation of a model’s predictive ability makes (Q)SAR analysis more transparent and increases confidence in the application of these predictions in a risk-assessment context. Using this approach, predictions for such chemicals obtained from four (Q)SAR models were successfully classified into high, medium and low confidence levels to better inform their use in decision-making.     

Evaluation and improvement of QSAR predictions of skin sensitization for pesticides

In vivo skin sensitization assays have to be provided by applicants to the competent authorities in the European Union for the approval of active substances (AS) in pesticides. This study aimed to test the practicability of in silico predictions for AS by freely available (Q)SAR tools to evaluate their use as a time- and cost-effective alternative to animal testing in the context of the 3R concept. Predictions of skin sensitization for 48 selected sensitizing and non-sensitizing AS by the software programs CAESAR, Toxtree, OECD (Q)SAR Toolbox, CASE Ultra, Leadscope and SciQSAR were collected and compared. Different data evaluation methodologies (score definition, mean, weighted mean, threshold score definition) were applied to optimize the predictions. The calculation methods were internally cross-validated and further validated with an additional validation set of 80 AS. Although the presented calculation methodologies are not suitable as a stand-alone method, this study has shown weaknesses and strengths of some prominent (Q)SAR programs and diverse combinatorial options in the prediction of skin sensitization by pesticidal AS. The present study will help to foster discussions on in silico alternatives to animal testing in the pesticide area.     

Integration of structure-activity relationship and artificial intelligence systems to improve in silico prediction of ames test mutagenicity

The Ames mutagenicity test in Salmonella typhimurium is a bacterial short-term in vitro assay aimed at detecting the mutagenicity caused by chemicals. Mutagenicity is considered as an early alert for carcinogenicity. After a number of decades, several (Q)SAR studies on this endpoint yielded enough evidence to make feasible the construction of reliable computational models for prediction of mutagenicity from the molecular structure of chemicals. In this study, we propose a combination of a fragment-based SAR model and an inductive database. The hybrid system was developed using a collection of 4337 chemicals (2401 mutagens and 1936 nonmutagens) and tested using 753 independent compounds (437 mutagens and 316 nonmutagens). The overall error of this system on the external test set compounds is 15% (sensitivity = 15%, specificity = 15%), which is quantitatively similar to the experimental error of Ames test data (average interlaboratory reproducibility determined by the National Toxicology Program). Moreover, each single prediction is provided with a specific confidence level. The results obtained give confidence that this system can be applied to support early and rapid evaluation of the level of mutagenicity concern.     

Improving (Q)SAR predictions by examining bias in the selection of compounds for experimental testing

ABSTRACT

Existing data on structures and biological activities are limited and distributed unevenly across distinct molecular targets and chemical compounds. The question arises if these data represent an unbiased sample of the general population of chemical-biological interactions. To answer this question, we analyzed ChEMBL data for 87,583 molecules tested against 919 protein targets using supervised and unsupervised approaches. Hierarchical clustering of the Murcko frameworks generated using Chemistry Development Toolkit showed that the available data form a big diffuse cloud without apparent structure. In contrast hereto, PASS-based classifiers allowed prediction whether the compound had been tested against the particular molecular target, despite whether it was active or not. Thus, one may conclude that the selection of chemical compounds for testing against specific targets is biased, probably due to the influence of prior knowledge. We assessed the possibility to improve (Q)SAR predictions using this fact: PASS prediction of the interaction with the particular target for compounds predicted as tested against the target has significantly higher accuracy than for those predicted as untested (average ROC AUC are about 0.87 and 0.75, respectively). Thus, considering the existing bias in the data of the training set may increase the performance of virtual screening.     

Global (Q)SARs for skin sensitisation–assessment against OECD principles‖

As part of a European Chemicals Bureau contract relating to the evaluation of (Q)SARs for toxicological endpoints of regulatory importance, we have reviewed and analysed (Q)SARs for skin sensitisation. Here we consider some recently published global (Q)SAR approaches against the OECD principles and present re-analysis of the data. Our analyses indicate that “statistical” (Q)SARs which aim to be global in their applicability tend to be insufficiently robust mechanistically, leading to an unacceptably high failure rate. Our conclusions are that, for skin sensitisation, the mechanistic chemistry is very important and consequently the best non-animal approach currently applicable to predict skin sensitisation potential is with the help of an expert system. This would assign compounds into mechanistic applicability domains and apply mechanism-based (Q)SARs specific for those domains and, very importantly, recognise when a compound is outside its range of competence. In such situations, it would call for human expert input supported by experimental chemistry studies as necessary.     

Progress and perspectives of quantitative structure-activity relationships used for ecological risk assessment of toxic organic compounds

Structure-activity relationship (SAR) and quantitative structure-activity relationship (QSAR), collec- tively referred to as (Q)SARs, play an important role in ecological risk assessment (ERA) of organic chemicals. (Q)SARs can fill the data gap for physical-chemical, environmental behavioral and ecotoxicological parameters of organic compounds; they can decrease experimental expenses and reduce the extent of experimental testing (especially animal testing); they can also be used to assess the uncertainty of the experimental data. With the development for several decades, (Q)SARs in envi- ronmental sciences show three features: application orientation, multidisciplinary integration, and in- telligence. Progress of (Q)SAR technology for ERA of toxic organic compounds, including endpoint selection and mathematic methods for establishing simple, transparent, easily interpretable and portable (Q)SAR models, is reviewed. The recent development on defining application domains and diagnosing outliers is summarized. Model characterization with respect to goodness-of-fit, stability and predictive power is specially presented. The purpose of the review is to promote the development of (Q)SARs orientated to ERA of organic chemicals.     

Decision tree SAR models for developmental toxicity based on an FDA/TERIS database

Humans are exposed to thousands of environmental chemicals for which no developmental toxicity information is available. Structure-activity relationships (SARs) are models that could be used to efficiently predict the biological activity of potential developmental toxicants. However, at this time, no adequate SAR models of developmental toxicity are available for risk assessment. In the present study, a new developmental database was compiled by combining toxicity information from the Teratogen Information System (TERIS) and the Food and Drug Administration (FDA) guidelines. We implemented a decision tree modeling procedure, using Classification and Regression Tree software and a model ensemble approach termed bagging. We then assessed the empirical distributions of the prediction accuracy measures of the single and ensemble-based models, achieved by repeating our modeling experiment many times by repeated random partitioning of the working database. The decision tree developmental SAR models exhibited modest prediction accuracy. Bagging tended to enhance the accuracy of prediction. Also, the model ensemble approach reduced the variability of prediction measures compared to the single model approach. Further research with data derived from animal species- and endpoint-specific components of an extended and refined FDA/TERIS database has the potential to derive SAR models that would be useful in the developmental risk assessment of the thousands of untested chemicals.     

Comparative PBT screening using (Q)SAR tools within REACH legislation

Small to medium sized enterprises (SMEs) in the EU are facing challenges due to the introduction of new legislation designed to protect consumers and the environment, REACH (Registration, Evaluation, Authorisation and Restriction of CHemicals). There can be high costs associated with implementing REACH because data on mammalian toxicity, environmental toxicity and environmental fate properties is required and if this data is obtained experimentally the cost is significant. These costs can be reduced if reliable quantitative structure–activity relationships ((Q)SAR) models are instead used to obtain the required information. In this paper we investigate how easily freely available (Q)SAR models can be applied for persistent, bioaccumulative and toxic (PBT) screening of 17 chemicals of interest to SMEs. In this study the PBT predictions obtained from the more user-friendly PBT Profiler and the Danish(Q)SAR database for the chemicals were compared with the results taken directly from the EPI Suite software. It was found that these widely used (Q)SAR databases might have some errors and examples are provided. It was concluded that extra care must be taken when considering the use of these databases for PBT screening. In addition, to increase the likelihood of a correct prediction, data estimates from various (Q)SAR models relevant to the PBT endpoints must be compared.     

Computational Toxicology and the Generation of Mechanistic Hypotheses: γ-Butyrolactone

Abstract

In this study, we use SAR approaches in an attempt to elucidate the action of γ-butyrolactone (GBL), an illicit drug and a dietary supplement, that can cause coma and deaths in humans while exhibiting low systemic toxicity towards rodents.

The lack of systemic toxicity of GBL and of its metabolite(s) was also predicted by validated SAR models. In fact using diverse SAR models, the only significant SAR prediction was that GBL had the potential for inhibiting human cytochrome P4502D6 (CYP2D6). However, inhibition of that isozyme is not necessarily associated with toxicity. It is suggested that GBL users also abuse other substances. When GBL inhibits CYP2D6 this may prevent the CYP2D6-mediated detoxification of other toxicants simultaneously consumed by the GBL user.     

Integrated approach to assess the domain of applicability of some commercial (Q)SAR models1

An integrated framework of data analysis has been proposed to systematically address the determination of the domain of applicability (DA) of some commercial Quantitative Structure Activity Relationship ((Q)SAR) models based on the structure of test chemicals. This framework forms one of the important steps in dealing with the growing concerns on reliability of model-based predictions on toxicity of chemicals specifically in the regulatory context. The present study uses some of the well-known mutagenicity and carcinogenicity models that are available within the Casetox (MultiCASE Inc.) and TOPKAT (Accelrys Software Inc.) programs. The approach enumerated in this paper employs chemoinformatics tools that facilitate comparisons of key structural features as well as application of cluster analysis techniques. The approach has been illustrated using a set of eleven chemical structures selected from the Canadian Domestic Substances List (DSL) that are not present in the model training sets, and the efficacy of the approach has also been assessed using seven chemicals with known toxicities. The methodologies presented here could help address the issue of DA of complex (Q)SAR models and at the same time, serve as useful tools for regulators to make a preliminary assessment of (Q)SAR based systems thereby helping the process of hazard-based regulatory assessments of chemicals.     

Computational Prediction of Inhibitors and Inducers of the Major Isoforms of Cytochrome P450

Human cytochrome P450 enzymes (CYPs) are heme-containing monooxygenases. This superfamily of drug-metabolizing enzymes is responsible for the metabolism of most drugs and other xenobiotics. The inhibition of CYPs may lead to drug–drug interactions and impair the biotransformation of drugs. CYP inducers may decrease the bioavailability and increase the clearance of drugs. Based on the freely available databases ChEMBL and PubChem, we have collected over 70,000 records containing the structures of inhibitors and inducers together with the IC50 values for the inhibitors of the five major human CYPs: 1A2, 3A4, 2D6, 2C9, and 2C19. Based on the collected data, we developed (Q)SAR models for predicting inhibitors and inducers of these CYPs using GUSAR and PASS software. The developed (Q)SAR models could be applied for assessment of the interaction of novel drug-like substances with the major human CYPs. The created (Q)SAR models demonstrated reasonable accuracy of prediction. They have been implemented in the web application P450-Analyzer that is freely available via the Internet.     

PASS-based prediction of metabolites detection in biological systems

ABSTRACT

Metabolite identification is an essential part of the drug discovery and development process. Experimental methods allow identifying metabolites and estimating their relative amount, but they require cost-intensive and time-consuming techniques. Computational methods for metabolite prediction are devoid of these shortcomings and may be applied at the early stage of drug discovery. In this study, we investigated the possibility of creating SAR models for the prediction of the qualitative metabolite yield (‘major’, ‘minor’, ”trace” and ”negligible”) depending on species and biological experimental systems. In addition, we have created models for prediction of xenobiotic excretion depending on its administration route for different species. The prediction is based on an algorithm of naïve Bayes classifier implemented in PASS software. The average accuracy of prediction was 0.91 for qualitative metabolite yield prediction and 0.89 for prediction of xenobiotic excretion. The created models were included as a component of MetaTox web application, which allows predicting the xenobiotic metabolism pathways (http://www.way2drug.com/mg).     

A stepwise approach for defining the applicability domain of SAR and QSAR models

A stepwise approach for determining the model applicability domain is proposed. Four stages are applied to account for the diversity and complexity of the current SAR/QSAR models, reflecting their mechanistic rationality (including metabolic activation of chemicals) and transparency. General parametric requirements are imposed in the first stage, specifying in the domain only those chemicals that fall in the range of variation of the physicochemical properties of the chemicals in the training set. The second stage defines the structural similarity between chemicals that are correctly predicted by the model. The structural neighborhood of atom-centered fragments is used to determine this similarity. The third stage in defining the domain is based on a mechanistic understanding of the modeled phenomenon. Here, the model domain combines the reliability of specific reactive groups hypothesized to cause the effect and the domain of explanatory variables determining the parametric requirements in order for functional groups to elicit their reactivity. Finally, the reliability of simulated metabolism (metabolites, pathways, and maps) is taken into account in assessing the reliability of predictions, if metabolic activation of chemicals is a part of the (Q)SAR model. Some of the stages of the proposed approach for defining the model domain can be eliminated depending on the availability and quality of the experimental data used to derive the model, the specificity of (Q)SARs, and the goals of their ultimate application. The performance of the proposed definition of the model domain is tested using several examples of (Q)SARs that have been externally validated, including models for predicting acute toxicity, skin sensitization, and biodegradation. The results clearly showed that credibility in predictions of QSAR models for chemicals belonging to their domain is much higher than for chemicals outside this domain.     

Assessment of applicability domain for multivariate counter-propagation artificial neural network predictive models by minimum Euclidean distance space analysis: A case study

Alongside the validation, the concept of applicability domain (AD) is probably one of the most important aspects which determine the quality as well as reliability of the established quantitative structure–activity relationship (QSAR) models. To date, a variety of approaches for AD estimation have been devised which can be applied to particular type of QSAR models and their practical utilization is extensively elaborated in the literature. The present study introduces a novel, simple, and effective distance-based method for estimation of the AD in case of developed and validated predictive counter-propagation artificial neural network (CP ANN) models through a proficient exploitation of the Euclidean distance (ED) metric in the structure-representation vector space. The performance of the method was evaluated and explained in a case study by using a pre-built and validated CP ANN model for prediction of the transport activity of the transmembrane protein bilitranslocase for a diverse set of compounds. The method was tested on two more datasets in order to confirm its performance for evaluation of the applicability domain in CP ANN models. The chemical compounds determined as potential outliers, i.e., outside of the CP ANN model AD, were confirmed in a comparative AD assessment by using the leverage approach. Moreover, the method offers a graphical depiction of the AD for fast and simple determination of the extreme points.     

In silico exploratory study using structure–activity relationship models and metabolic information for prediction of mutagenicity based on the Ames test and rodent micronucleus assay

The mutagenic potential of chemicals is a cause of growing concern, due to the possible impact on human health. In this paper we have developed a knowledge-based approach, combining information from structure–activity relationship (SAR) and metabolic triggers generated from the metabolic fate of chemicals in biological systems for prediction of mutagenicity in vitro based on the Ames test and in vivo based on the rodent micronucleus assay. In the first part of the work, a model was developed, which comprises newly generated SAR rules and a set of metabolic triggers. These SAR rules and metabolic triggers were further externally validated to predict mutagenicity in vitro, with metabolic triggers being used only to predict mutagenicity of chemicals, which were predicted unknown, by SARpy. Hence, this model has a higher accuracy than the SAR model, with an accuracy of 89% for the training set and 75% for the external validation set. Subsequently, the results of the second part of this work enlist a set of metabolic triggers for prediction of mutagenicity in vivo, based on the rodent micronucleus assay. Finally, the results of the third part enlist a list of metabolic triggers to find similarities and differences in the mutagenic response of chemicals in vitro and in vivo.     

A Novel Electron-Conformational Approach to Molecular Modeling for QSAR by Identification of Pharmacophore and Anti-Pharmacophore Shielding

Abstract

A novel method of pharmacophore identification and activity prediction in structure-activity (structure-property) relationships is worked out as an essential extension and improvement of previous publications. In this method each conformation of the molecular systems in the training set of the SAR problem is presented by both electronic structure and geometry parameters arranged in a matrix form. Multiple comparisons of these matrices for the active and inactive compounds allows one to separate a smaller number of matrix elements that are common for all the active compounds and are not present in the same arrangement in the inactive ones. This submatrix of activity represents the pharmacophore (Pha).

By introducing the Anti-Pharmacophore Shielding (APS) defined as molecular groups and competing charges outside the Pha that hinder the proper docking of the Pha with the bioreceptor, the procedure of Pha identification is essentially reduced to the treatment of a smaller number of simplest in structure most active and inactive compounds. A simple empirical scheme is suggested to estimate the APS numerically, while the contributions of different conformations of the same compound are taken into account by means of Boltzmann distribution. This enables us to make approximate quantitative predictions of activities.

In application to rice blast activity we reached an approximately 100% (within experimental error) prediction probability of the activity qualitatively (yes, no), and with r ² = 70% quantitatively.