Ensemble QSAR Modeling to Predict Multispecies Fish Toxicity Lethal Concentrations and Points of Departure

Transferring knowledge across aquatic species via clustering techniques to unravel patterns of pesticide toxicity

In silico modelling takes the advantage of accelerating ecotoxicological assessments on hazardous chemicals without conducting risky in vivo experiments under ethic regulation. To date, the prevailing strategy of one model for one species cannot be well generalized to multi-species modelling. In this work, we propose a new strategy of one model for multiple species to facilitate knowledge transfer across aquatic species. The available lethal concentration values of 4952 pesticides on 651 fish species are aggregated into one toxicity response matrix, purely through which we attempt to unravel fish toxicosis-phylogenesis relationships and pesticide toxicity-structure relationships via clustering techniques including non-negative matrix factorization (NMF) and hierarchical clustering. The clustering results suggest that (1) close NMF weights indicate close species-toxicosis and pesticide-toxicity profiles; (2) and that species toxicosis patterns are related with species phylogenetic relationships; (3) and that close pesticide-toxicity profiles indicate similar atom-pair structural fingerprints. These environmental, chemical and biological insights can be used as expert knowledge for environmentalists to manually gain knowledge about untested species/pesticides from tested species/pesticides, and meanwhile provide support for us to build in silico models from species phylogenetic and pesticide structural points of view. Besides unravelling the mechanisms behind toxicity response, we also adopt stratified cross validation and external test to validate the reliability of using NMF to predict missing toxicity values. Independent test on external data shows that NMF achieves 0.8404-0.9397 R2 on four fish species. In the context of toxicity prediction, non-negative matrix factorization can be viewed as a model based on quantitative activity-activity relationships (QAAR), and provides an alternative approach of inferring toxicity values on untested species from tested species.

Molecular similarity in chemical informatics and predictive toxicity modeling: from quantitative read-across (q-RA) to quantitative read-across structure-activity relationship (q-RASAR) with the application of machine learning

This article aims to provide a comprehensive critical, yet readable, review of general interest to the chemistry community on molecular similarity as applied to chemical informatics and predictive modeling with a special focus on read-across (RA) and read-across structure-activity relationships (RASAR). Molecular similarity-based computational tools, such as quantitative structure-activity relationships (QSARs) and RA, are routinely used to fill the data gaps for a wide range of properties including toxicity endpoints for regulatory purposes. This review will explore the background of RA starting from how structural information has been used through to how other similarity contexts such as physicochemical, absorption, distribution, metabolism, and elimination (ADME) properties, and biological aspects are being characterized. More recent developments of RA's integration with QSAR have resulted in the emergence of novel models such as ToxRead, generalized read-across (GenRA), and quantitative RASAR (q-RASAR). Conventional QSAR techniques have been excluded from this review except where necessary for context.

Prediction of chronic toxicity of pharmaceuticals in Daphnia magna by combining ortholog prediction, pharmacological effects, and quantitative structure-activity relationship

To develop a method for predicting chronic toxicity of pharmaceuticals in Daphnia, we investigated the feasibility of combining the presence of drug-target orthologs in Daphnia magna, classification based on pharmacological effects, and ecotoxicity quantitative structure-activity relationship (QSAR) prediction. We established datasets on the chronic toxicity of pharmaceuticals in Daphnia, including information on therapeutic categories, target proteins, and the presence or absence of drug-target orthologs in D. magna, using literature and databases. Chronic toxicity was predicted using ecotoxicity prediction QSAR (Ecological Structure Activity Relationship and Kashinhou Tool for Ecotoxicity), and the differences between the predicted and measured values and the presence or absence of drug-target orthologs were examined. For pharmaceuticals without drug-target orthologs in D. magna or without expected specific actions, the ecotoxicity prediction QSAR analysis yielded acceptable predictions of the chronic toxicity of pharmaceuticals. In addition, a workflow model to assess the chronic toxicity of pharmaceuticals in Daphnia was proposed based on these evaluations and verified using an additional dataset. The addition of biological aspects such as drug-target orthologs and pharmacological effects would support the use of QSARs for predicting the chronic toxicity of pharmaceuticals in Daphnia.

ECOTOXr: An R Package for Reproducible and Transparent Retrieval of Data from EPA's ECOTOX Database

The US EPA ECOTOX database provides key ecotoxicological data that are crucial in environmental risk assessment. It can be used for computational predictions of toxicity or indications of hazard in a wide range of situations. There is no standardised or formalised method for extracting and subsetting data from the database for these purposes. Consequently, results to such meta-analyses are difficult to reproduce. The present study introduces the software package ECOTOXr, which provides the means to formalise data retrieval from the ECOTOX database in the R scripting language. Three cases are presented to evaluate the performance of the package in relation to earlier data extractions and searches on the website. These cases demonstrate that the package can reproduce data sets relatively well. Furthermore, they illustrate how future studies can further improve traceability and reproducibility by applying the package and adhering to some simple guidelines. This contributes to the FAIR principles, credibility and acceptance of research that uses data from the ECOTOX database.

Photo-transformation of ofloxacin in natural aquatic conditions: Impact of indirect photolysis on the product profile and transformation mechanism

The natural phototransformation of organic pollutants in the environment depends on several water constituents, including inorganic ions, humic substances, and pH. However, the literature information concerning the influence of various water components on the amount of phototransformation and their impact on the development of various transformation products (TPs) is minimal. This study investigated the phototransformation of ofloxacin (OFL), a fluoroquinolone antibiotic, in the presence of various water components such as cations (K+, Na+, Ca2+, NH4+, Mg2+), anions (NO3-, SO42-, HCO3-, CO32-, PO43-), pH, and humic substances when exposed to natural sunlight. The study reveals that neutral pH levels (0.39374 min⁻1) enhance the phototransformation of OFL in aquatic environments. Carbonate, among anions, shows the highest rate constant (2.89966 min⁻1), significantly influencing OFL phototransformation, while all anions exhibit a notable impact. In aquatic environments, indirect phototransformation of OFL, driven by increased reactive oxygen species, expedites light-induced reactions, potentially enhancing OFL phototransformation. A clear difference was visible in the type of transformation products (TPs) formed during direct and indirect photolysis. The impact of indirect photolysis in the product profile was evaluated by examining the unique properties of TPs in direct and indirect photolysis. The primary transformation products were generated by oxidation and cleavage processes directed towards the ofloxacin piperazinyl, oxazine, and carboxyl groups. The toxicity assessment of TPs derived from OFL revealed that among the 26 identified TPs, TP3 (demethylated product), TP7 and TP8 (decarboxylated products), and TP15 (piperazine ring cleaved product) could potentially have some toxicological effects. These findings suggest that the phototransformation of OFL in the presence of various water components is necessary when assessing this antibiotic's environmental fate.

Oxygen vacancy based WO3/SnO2-x promote electrochemical H2O2 accumulation by two-electron water oxidation reaction and toxic uniform dimethylhydrazine degradation

The key to constructing an anodic electro-Fenton system hinges on two pivotal criteria: enhancing the catalyst activity and selectivity in water oxidation reaction (WOR), while simultaneously inhibiting the decomposition of hydrogen peroxide (H2O2) which is on-site electrosynthesized at the anode. To address the issues, we synthesized novel WO3/SnO2-x electrocatalysts, enriched with oxygen vacancies, capitalize on the combined activity and selectivity advantages of both WO3 and SnO2-x for the two-electron pathway electrocatalytic production of H2O2. Moreover, the introduction of oxygen vacancies plays a critical role in impeding the decomposition of H2O2. This innovative design ensures that the Faraday efficiency and yield of H2O2 are maintained at over 80 %, with a noteworthy production rate of 0.2 mmol h-1 cm-2. We constructed a novel electro-Fenton system that operates using only H2O as its feedstock and applied it to treat highly toxic uniform dimethylhydrazine (UDMH) from rocket launch effluent. Our experiments revealed a substantial total organic carbon (TOC) removal, achieving approximately 90 % after 120 mins of treatment. Additionally, the toxicity of N-nitrosodimethylamine (NDMA), a byproduct of great concern, was shown to be effectively mitigated, as evidenced by acute toxicity evaluations using zebrafish embryos. The degradation mechanism of UDMH is predominantly characterized by the advanced oxidative action of H2O2 and hydroxyl radicals, as well as by complex electron transfer processes that warrant further investigation.

Hybrid radical coupling during MnO2-mediated transformation of a mixture of organic UV filters: Chemistry and toxicity assessment

Manganese oxide (MnO2) is one of the most abundant metal oxides, and it is renowned for its ability to degrade various phenolic micropollutants. However, under MnO2-mediated transformation, BP-3 transforms into 12 different radical-coupled transformation products (TPs) out of 15 identified TPs. These radical-coupled TPs are reported with adverse environmental impacts. This study explored the effects of MnO2 on organic UV filter mixtures and different water constituents (i.e., bicarbonate ion (HCO3-), humic acid (HA) and halide ions) in terms of degradation efficiency and transformation chemistry. When a mixture of benzophenone-3 (BP-3) and avobenzone (AVO) underwent transformation by MnO2, hybrid radical-coupled TPs derived from both organic UV filters were generated. These hybrid radical-coupled TPs were evaluated by an in silico prediction tool and Vibrio fischeri bioluminescence inhibition assay (VFBIA). Results showed that these TPs were potentially toxic to aquatic organisms, even more so than their parent compounds. The higher the concentration of HCO3-, HA, chloride ion (Cl-) and bromide ion (Br-), the greater the reduction in the efficiencies of degrading BP-3 and AVO. Contrastingly, in the presence of iodide ion (I-), degradation efficiencies of BP-3 and AVO were enhanced; however, iodinated TPs and iodinated radical-coupled TPs were formed, with questionable toxicity. This study has revealed the environmental risks of hybrid radical-coupled TPs, iodinated TPs and iodinated radical-coupled TPs when the organic UV filters BP-3 and AVO are transformed by MnO2.

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.

Multimodal Model to Predict Tissue-to-Blood Partition Coefficients of Chemicals in Mammals and Fish

Tissue-to-blood partition coefficients (Ptb) are key parameters for assessing toxicokinetics of xenobiotics in organisms, yet their experimental data were lacking. Experimental methods for measuring Ptb values are inefficient, underscoring the urgent need for prediction models. However, most existing models failed to fully exploit Ptb data from diverse sources, and their applicability domain (AD) was limited. The current study developed a multimodal model capable of processing and integrating textual (categorical features) and numerical information (molecular descriptors/fingerprints) to simultaneously predict Ptb values across various species, tissues, blood matrices, and measurement methods. Artificial neural network algorithms with embedding layers were used for the multimodal modeling. The corresponding unimodal models were developed for comparison. Results showed that the multimodal model outperformed unimodal models. To enhance the reliability of the model, a method considering categorical features, weighted molecular similarity density, and weighted inconsistency in molecular activities of structure-activity landscapes was used to characterize the AD. The model constrained by the AD exhibited better prediction accuracy for the validation set, with the determination coefficient, root mean-square error, and mean absolute error being 0.843, 0.276, and 0.213 log units, respectively. The multimodal model coupled with the AD characterization can serve as an efficient tool for internal exposure assessment of chemicals.

Hazardous radical-coupled transformation products of benzophenone-3 formed during manganese dioxide treatment

Radical-coupled transformation products (TPs) have been identified as the byproducts of various transformation processes, including both natural attenuation and artificial treatments, of phenolic micropollutants. Benzophenone-3 (BP-3), an organic UV filter of emerging concern, has been previously reported with ubiquitous occurrence in the natural environment and water bodies. Current research has demonstrated how TPs are formed from BP-3 when it is treated with manganese oxide (MnO2). The ecological and toxicological risks of these TPs have also been assessed. Polymerization of BP-3 through radical coupling was observed as the major pathway by which BP-3 is transformed when treated with MnO2. These radical-coupled TPs haven't shown further degradation after formation, suggesting their potential persistence once occurred in the environment. In silico experiments predict the radical-coupled TPs will increase in mobility, persistence and ecotoxicity. If true, they also represent an ever-increasing threat to the environment, ecosystems and, most immediately, aquatic living organisms. In addition, radical-coupled TPs produced by MnO2 transformation of BP-3 have shown escalated estrogenic activity compared to the parent compound. This suggests that radical coupling amplifies the toxicological impacts of parent compound. These results provide strong evidence that radical-coupled TPs with larger molecular sizes are having potential adverse impacts on the ecosystem and biota.

The Bigger Fish: A Comparison of Meta-Learning QSAR Models on Low-Resourced Aquatic Toxicity Regression Tasks

Toxicological information as needed for risk assessments of chemical compounds is often sparse. Unfortunately, gathering new toxicological information experimentally often involves animal testing. Simulated alternatives, e.g., quantitative structure-activity relationship (QSAR) models, are preferred to infer the toxicity of new compounds. Aquatic toxicity data collections consist of many related tasks─each predicting the toxicity of new compounds on a given species. Since many of these tasks are inherently low-resource, i.e., involve few associated compounds, this is challenging. Meta-learning is a subfield of artificial intelligence that can lead to more accurate models by enabling the utilization of information across tasks. In our work, we benchmark various state-of-the-art meta-learning techniques for building QSAR models, focusing on knowledge sharing between species. Specifically, we employ and compare transformational machine learning, model-agnostic meta-learning, fine-tuning, and multi-task models. Our experiments show that established knowledge-sharing techniques outperform single-task approaches. We recommend the use of multi-task random forest models for aquatic toxicity modeling, which matched or exceeded the performance of other approaches and robustly produced good results in the low-resource settings we studied. This model functions on a species level, predicting toxicity for multiple species across various phyla, with flexible exposure duration and on a large chemical applicability domain.

What is the ecotoxicity of a given chemical for a given aquatic species? Predicting interactions between species and chemicals using recommender system techniques

Ecotoxicological safety assessment of chemicals requires toxicity data on multiple species, despite the general desire of minimizing animal testing. Predictive models, specifically machine learning (ML) methods, are one of the tools capable of solving this apparent contradiction as they allow to generalize toxicity patterns across chemicals and species. However, despite the availability of large public toxicity datasets, the data is highly sparse, complicating model development. The aim of this study is to provide insights into how ML can predict toxicity using a large but sparse dataset. We developed models to predict LC50-values, based on experimental LC50-data covering 2431 organic chemicals and 1506 aquatic species from the ECOTOX-database. Several well-known ML techniques were evaluated and a new ML model was developed, inspired by recommender systems. This new model involves a simple linear model that learns low-rank interactions between species and chemicals using factorization machines. We evaluated the predictive performances of the developed models based on two validation settings: 1) predicting unseen chemical-species pairs, and 2) predicting unseen chemicals. The results of this study show that ML models can accurately predict LC50-values in both validation settings. Moreover, we show that the novel factorization machine approach can match well-tuned, complex, ML approaches.

Physiological variables in machine learning QSARs allow for both cross-chemical and cross-species predictions

A major challenge in ecological risk assessment is estimating chemical-induced effects across taxa without species-specific testing. Where ecotoxicological data may be more challenging to gather, information on species physiology is more available for a broad range of taxa. Physiology is known to drive species sensitivity but understanding about the relative contribution of specific underlying processes is still elusive. Consequently, there remains a need to understand which physiological processes lead to differences in species sensitivity. The objective of our study was to utilize existing knowledge about organismal physiology to both understand and predict differences in species sensitivity. Machine learning models were trained to predict chemical- and species-specific endpoints as a function of both chemical fingerprints/descriptors and physiological properties represented by dynamic energy budget (DEB) parameters. We found that random forest models were able to predict chemical- and species-specific endpoints, and that DEB parameters were relatively important in the models, particularly for invertebrates. Our approach illuminates how physiological properties may drive species sensitivity, which will allow more realistic predictions of effects across species without the need for additional animal testing.

Quantitative multi-species toxicity modeling: Does a multi-species, machine learning model provide better performance than a single-species model for the evaluation of acute aquatic toxicity by organic pollutants?

The toxicological profile of any chemical is defined by multiple endpoints and testing procedures, including representative test species from different trophic levels. While computer-aided methods play an increasingly important role in supporting ecotoxicology research and chemical hazard assessment, most of the recently developed machine learning models are directed towards a single, specific endpoint. To overcome this limitation and accelerate the process of identifying potentially hazardous environmental pollutants, we are introducing an effective approach for quantitative, multi-species modeling. The proposed approach is based on canonical correlation analysis that finds a pair(s) of uncorrelated, linear combinations of the original variables that best defines the overall variability within and between multiple biological responses and predictor variables. Its effectiveness was confirmed by the machine learning model for estimating acute toxicity of diverse organic pollutants in aquatic species from three trophic levels: algae (Pseudokirchneriella subcapitata), daphnia (Daphnia magna), and fish (Oryzias latipes). The multi-species model achieved a favorable predictive performance that were in line with predictive models derived for the aquatic organisms individually. The chemical bioavailability and reactivity parameters (n-octanol/water partition coefficient, chemical potential, and molecular size and volume) were important to accurately predict acute ecotoxicity to the three aquatic organisms. To facilitate the use of this approach, an open-source, Python-based script, named qMTM (quantitative Multi-species Toxicity Modeling) has been provided.

Prior Knowledge for Predictive Modeling: The Case of Acute Aquatic Toxicity

Early assessment of the potential impact of chemicals on health and the environment requires toxicological properties of the molecules. Predictive modeling is often used to estimate the property values in silico from pre-existing experimental data, which is often scarce and uncertain. One of the ways to advance the predictive modeling procedure might be the use of knowledge existing in the field. Scientific publications contain a vast amount of knowledge. However, the amount of manual work required to process the enormous volumes of information gathered in scientific articles might hinder its utilization. This work explores the opportunity of semiautomated knowledge extraction from scientific papers and investigates a few potential ways of its use for predictive modeling. The knowledge extraction and predictive modeling are applied to the field of acute aquatic toxicity. Acute aquatic toxicity is an important parameter of the safety assessment of chemicals. The extensive amount of diverse information existing in the field makes acute aquatic toxicity an attractive area for investigation of knowledge use for predictive modeling. The work demonstrates that the knowledge collection and classification procedure could be useful in hybrid modeling studies concerning the model and predictor selection, addressing data gaps, and evaluation of models' performance.

Linking electron ionization mass spectra of organic chemicals to toxicity endpoints through machine learning and experimentation

Quantitative structure-activity relationship (QSAR) modeling has been widely used to predict the potential harm of chemicals, in which the prediction heavily relies on the accurate annotation of chemical structures. However, it is difficult to determine the accurate structure of an unknown compound in many cases, such as in complex water environments. Here, we solved the above problem by linking electron ionization mass spectra (EI-MS) of organic chemicals to toxicity endpoints through various machine learning methods. The proposed method was verified by predicting 50% growth inhibition of Tetrahymena pyriformis (T. pyriformis) and liver toxicity. The optimal model performance obtained an R² > 0.7 or balanced accuracy > 0.72 for both the training set and test set. External experimentation further verified the application potential of our proposed method in the toxicity prediction of unknown chemicals. Feature importance analysis allowed us to identify critical spectral features that were responsible for chemical-induced toxicity. Our approach has the potential for toxicity prediction in such fields that it is difficult to determine accurate chemical structures.

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, i.e. toxic (LC₅₀ < 10 ppm) 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 multi-task GCN showed no advantages over the conventional machine learning methods. The data and models are available on GitHub (https://github.com/ChemPredict/ChemicalAquaticToxicity).

In silico prediction models for thyroid peroxidase inhibitors and their application to synthetic flavors

Systematic toxicity tests are often waived for the synthetic flavors as they are added in a very small amount in foods. However, their safety for some endpoints such as endocrine disruption should be concerned as they are likely to be active in low levels. In this case, structure–activity-relationship (SAR) models are good alternatives. In this study, therefore, binary, ternary, and quaternary prediction models were designed using simple or complex machine-learning methods. Overall, hard-voting classifiers outperformed other methods. The test scores for the best binary, ternary, and quaternary models were 0.6635, 0.5083, and 0.5217, respectively. Along with model development, some substructures including primary aromatic amine, (enol)ether, phenol, heterocyclic sulfur, and heterocyclic nitrogen, dominantly occurred in the most highly active compounds. The best predicting models were applied to synthetic flavors, and 22 agents appeared to have a strong inhibitory potential towards TPO activities.

Unswerving modeling of hepatotoxicity of cadmium containing quantum dots using amalgamation of quasiSMILES, index of ideality of correlation, and consensus modeling

Liver toxicity of quantum dots varies with size, concentration, and other structural as well as experimental parameters. For modeling hepatotoxicity, the eclectic data associated with cadmium containing quantum dots have been used in the creation of quasiSMILES for their representation. The core diameter is normalized for wider applicability and the index of the ideality of correlation is applied to construct the better quantitative features toxicity relationship models. Total eight splits are created and the best model is obtained through split 1 with better prediction criteria of validation set objects. The values of all statistical criteria used in the quality determination of a QSAR model are within the specified range for all the eight toxicity models developed here. Factors like TGA ligand and 0.6-0.7 nm diameter are favorable for liver toxicity while L-cysteine ligand and 0.5-0.6 nm core diameter are helpful in the reduction of toxicity. Further, the intelligent consensus modeling process forms a total of 40 individual and 20 consensus models and the best individual and consensus models are 'Good' in MAE-based criteria. The consensus modeling enhances the prediction ability as well as the accuracy of the developed models and increases the applicability space of the built models for hepatotoxicity prediction of quantum dots.

Predicting formation of haloacetic acids by chlorination of organic compounds using machine-learning-assisted quantitative structure-activity relationships

The presence of disinfection byproducts (DBPs) in drinking water is a major public health concern, and an effective strategy to limit the formation of these DBPs is to prevent their precursors. In silico prediction from chemical structure would allow rapid identification of precursors and could be used as a prescreening tool to prioritize testing. We present models using machine learning algorithms (i.e., support vector regressor, random forest regressor, and multilayer perceptron regressor) and chemical descriptors as features to predict the formation of haloacetic acids (HAAs). A robust model with good predictivity (i.e., leave-one-out cross-validated Q² > 0.5) to predict the formation of trichloroacetic acid (TCAA) was developed using a random forest regressor. The number of aromatic bonds, hydrophilicity, and electrotopological descriptors related to electrostatic interactions and the atomic distribution of electronegativity were identified as important predictors of TCAA formation potentials (FPs). However, the prediction of dichloroacetic acid was less accurate, which is congruent with the presence of different types of precursors exhibiting distinct mechanisms. This study demonstrates that nonlinear combinations of general chemical descriptors can adequately estimate HAAFPs, and we hope that our study can be used to predict precursors of other disinfection byproducts based on chemical structures using a similar workflow.

Comparison of seven in silico tools for evaluating of daphnia and fish acute toxicity: case study on Chinese Priority Controlled Chemicals and new chemicals

BACKGROUND

A number of predictive models for aquatic toxicity are available, however, the accuracy and extent of easy to use of these in silico tools in risk assessment still need further studied. This study evaluated the performance of seven in silico tools to daphnia and fish: ECOSAR, T.E.S.T., Danish QSAR Database, VEGA, KATE, Read Across and Trent Analysis. 37 Priority Controlled Chemicals in China (PCCs) and 92 New Chemicals (NCs) were used as validation dataset.

RESULTS

In the quantitative evaluation to PCCs with the criteria of 10-fold difference between experimental value and estimated value, the accuracies of VEGA is the highest among all of the models, both in prediction of daphnia and fish acute toxicity, with accuracies of 100% and 90% after considering AD, respectively. The performance of KATE, ECOSAR and T.E.S.T. is similar, with accuracies are slightly lower than VEGA. The accuracy of Danish Q.D. is the lowest among the above tools with which QSAR is the main mechanism. The performance of Read Across and Trent Analysis is lowest among all of the tested in silico tools. The predictive ability of models to NCs was lower than that of PCCs possibly because never appeared in training set of the models, and ECOSAR perform best than other in silico tools.

CONCLUSION

QSAR based in silico tools had the greater prediction accuracy than category approach (Read Across and Trent Analysis) in predicting the acute toxicity of daphnia and fish. Category approach (Read Across and Trent Analysis) requires expert knowledge to be utilized effectively. ECOSAR performs well in both PCCs and NCs, and the application shoud be promoted in both risk assessment and priority activities. We suggest that distribution of multiple data and water solubility should be considered when developing in silico models. Both more intelligent in silico tools and testing are necessary to identify hazards of Chemicals.

Consensus QSAR models estimating acute toxicity to aquatic organisms from different trophic levels: algae, Daphnia and fish

We report new consensus models estimating acute toxicity for algae, Daphnia and fish endpoints. We assembled a large collection of 3680 public unique compounds annotated by, at least, one experimental value for the given endpoint. Support Vector Machine models were internally and externally validated following the OECD principles. Reasonable predictive performances were achieved (RMSE_ext = 0.56-0.78) which are in line with those of state-of-the-art models. The known structural alerts are compared with analysis of the atomic contributions to these models obtained using the ISIDA/ColorAtom utility. A benchmarking against existing tools has been carried out on a set of compounds considered more representative and relevant for the chemical space of the current chemical industry. Our model scored one of the best accuracy and data coverage. Nevertheless, industrial data performances were noticeably lower than those on public data, indicating that existing models fail to meet the industrial needs. Thus, final models were updated with the inclusion of new industrial compounds, extending the applicability domain and relevance for application in an industrial context. Generated models and collected public data are made freely available.