Computational Toxicology最新文献

Volatile sulfur compounds (VSCs) are highly volatile and most frequently associated with oral malodor. The odor quality is associated with the size and shape of the molecule along with stability, hydrogen bonding, extended d-shell electronic behavior, and complicity of d-shell bonding. Chemical reactivity descriptors of VSCs, such as chemical hardness (η), softness (σ), chemical potential (μ), electrophilic index (ω), and electronegativity (χ) are computed at B₃LYP/Aug-cc-PVTZ level of theory from the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) in the light of Koopmans’ approximation. Furthermore, the global reactivity parameters are evaluated from the vertical ionization potential (IP) and electron affinity (EA) to support the results of Koopmans’ theorem. These reactivity parameters offer a quantitative measure of the electronic structure and chemical properties of VSCs, offering insights into their stability, reactivity, and interaction with other molecules. A Python-based application is provided for the rapid calculation of these parameters (GitHub: Link).

Read-across, wherein information from a data-rich chemical is used to make a prediction for a similar chemical that lacks the relevant data, is increasingly being accepted as an alternative to animal testing. Identifying chemicals that can be considered as similar (analogues) is crucial to the process. Two resources have been developed previously to address the issue of analogue selection and facilitate physiologically-based kinetic (PBK) model development, using read-across. Chemical-specific PBK models, available in the literature, were collated to form a PBK model dataset (PMD) of over 7,500 models. A KNIME workflow was created to accompany this PMD that can aid the selection of appropriate chemical analogues from chemicals within this dataset (i.e. chemicals that are similar to a target of interest and are known to have an existing PBK model). Information from the PBK model for the source chemical can then be used in a read-across approach to inform the development of a new PBK model for the target. The application of these resources is tested here using two case studies (i) for the drug atenolol and (ii) for the plant protection product, flumioxazin. New PBK models were constructed for these two target chemicals using data obtained from source chemicals, identified by the workflow as being similar (analogues). In each case, the published PBK model for the source chemical was initially reproduced, as accurately as possible, before being adapted and used as a template for the target chemical. The performance of the new PBK models was assessed by comparing simulation outputs to existing data on key kinetic properties for the targets. The results demonstrate that a read-across approach can be successfully applied to develop new PBK models for data-poor chemicals, thus enabling their deployment during early-stage risk assessment. This assists prediction of internal exposure whilst reducing reliance on animal testing.

The Toxicological Prioritization Index (ToxPi) is a visual analysis and decision support tool for dimension reduction and visualization of high throughput, multi-dimensional feature data. ToxPi was originally developed for assessing the relative toxicity of multiple chemicals or stressors by synthesizing complex toxicological data to provide a single comprehensive view of the potential health effects. It continues to be used for profiling chemicals and has since been applied to other types of “sample” entities, including geospatial (e.g. county-level Covid-19 risk and sites of historical PFAS exposure) and other profiling applications. For any set of features (data collected on a set of sample entities), ToxPi integrates the data into a set of weighted slices that provide a visual profile and a score metric for comparison. This scoring system is highly dependent on user-provided feature weights, yet users often lack knowledge of how to define these feature weights. Common methods for predicting feature weights are generally unusable due to inappropriate statistical assumptions and lack of global distributional expectation. However, users often have an inherent understanding of expected results for a small subset of samples. For example, in chemical toxicity, prior knowledge can often place subsets of chemicals into categories of low, moderate or high toxicity (reference chemicals). Ordinal regression can be used to predict weights based on these response levels that are applicable to the entire feature set, analogous to using positive and negative controls to contextualize an empirical distribution. We propose a semi-supervised method utilizing ordinal regression to predict a set of feature weights that produces the best fit for the known response (“reference”) data and subsequently fine-tunes the weights via a customized genetic algorithm. We conduct a simulation study to show when this method can improve the results of ordinal regression, allowing for accurate feature weight prediction and sample ranking in scenarios with minimal response data. To ground-truth the guided weight optimization, we test this method on published data to build a ToxPi model for comparison against expert-knowledge-driven weight assignments.

Read-across refers to the process by which information from one (source) chemical is used to infer information about another similar (target) chemical. This method can be used to fill data gaps and so inform safety assessment where data are lacking for chemicals of interest. As one chemical cannot be considered as absolutely similar to another, only similar with respect to a given property, it is essential to justify the selection of similar chemicals (analogues) for the purposes of read-across. A previously created dataset of available physiologically-based kinetic (PBK) models (referred to as the PBK modelling dataset or PMD) was used in the development of a KNIME workflow. KNIME is a freely-available, open-source analytics platform that allows users to create workflows to analyse and visualise data. The KNIME workflow described here was designed to identify chemical analogues with a corresponding model in the PMD. The PMD combined with the KWAAS enables PBK model information from source chemical(s) to be used in a read-across approach to help develop new PBK models for target chemicals. This KNIME workflow was applied to six chemicals, representing different types of chemical classes (drugs, cosmetics, botanicals, industrial chemicals, pesticides, and food additives) to assess its applicability across various industries. Information acquired from these PBK models can be used to support safety assessment of chemicals and reduce reliance on animal testing.

A molecular docking investigation was conducted to study the interaction between permethrin (PMT), a commonly used pyrethroid insecticide, known for its toxic effects on various organisms, including insects, aquatic life, and mammals, including humans with hemoglobin (HB). To assess its potential binding with the HB target, molecular docking simulations were conducted using different software. Each software has unique algorithms and scoring methods. Employing multiple tools helped us confirm and understand the interaction better. The results indicated high binding strengths across the various docking web servers. The PMT-HB complexation was largely stabilized via the hydrophobic interactions and Van der Waals forces. Also, PMT exhibited binding at a significant distance from the heme, indicating that it does not interfere with the essential biological function of HB, which is the binding of oxygen. In addition, the analysis of toxicological parameters revealed that PMT possesses the ability to induce acute oral and dermal toxicity.

A toxicokinetic modeling approach was used to study the transfer of 7 polychlorinated dibenzo-p-dioxins (PCDDs), 10 dibenzofurans (PCDFs), 12 dioxin-like polychlorinated biphenyls (dl-PCB) and 3 non-dioxin like (ndl) PCBs in dairy cows. The model describes the concentration–time profile of each congener in milk and blood of high-yielding dairy cows. It was parametrized using an in-house transfer study with 3 cows exposed to a defined synthetic congener mixture for two dosing periods, as well as 3 control cows to account for background exposure. The first dosing was administered during negative energy balance (NEB) after calving, and the second during positive energy balance (PEB) in late lactation. Results include extrapolated steady-state transfer rates and elimination half-lives, many of which have never been reported before. Transfer rates (TRs) were significantly higher during the NEB by a median of 27%, likely due to an increase in non-milk elimination during PEB. The difference draws attention to the influence of the metabolic state of food-producing animals in risk assessment. Comparison of the TRs derived here with those reported in the literature showed that they were, in median, 43% higher in the NEB phase and 16% higher in the PEB phase probably because we report TRs in steady-state unlike most literature sources.

Aromatase is a crucial enzyme in the aromatization process, which catalyzes the conversion of androgenic steroids to estrogens. Aromatase dysregulation, as well as elevated estrogen levels, have been linked to a variety of malignancies, including breast cancer. Herein, we present the results of the optimization of Xanthones employing density functional theory (DFT) using the B3LYP/6-311G+(d, p) basis set to determine their frontier molecular orbitals, Mulliken charges, and chemical reactivity descriptors. According to the DFT results, Erythrommone has the smallest HOMO-LUMO gap (3.85 Kcal/mol), as well as the greatest electrophilicity index (5.19) and basicity (4.47). Xanthones and their derivatives were docked into the active site cavity of CYP450 to examine their structure-based inhibitory effect. The docking simulation studies predicted that Erythrommone has the lowest binding energy (-7.43 Kcal/mol), which is consistent with the DFT calculations and may function as a powerful CYP450 inhibitor equivalent to its known inhibitor, Exemestane, which has a binding affinity of −8.13 Kcal/mol. The high binding affinity of Xanthones was linked to the existence of hydrogen bonds as well as various hydrophobic interactions between the ligand and the receptor's essential amino acid residues. The findings demonstrated that Xanthones are more powerful inhibitors of the Aromatase enzyme than the recognized inhibitor Exemestane.

Established cytotoxicity assays are commonly used for assessing the hepatotoxic risk of a compound. The addition of gene expression measurements from high-dimensional RNAseq experiments offers the potential for improved classification. However, it is generally not clear how best to summarize the high-dimensional gene measurements into meaningful variables. We propose several intuitive methods for dimension reduction of gene expression measurements toward interpretable variables and explore their relevance in predicting hepatotoxicity, using a dataset with 60 compounds.

Different advanced statistical learning algorithms are evaluated as classification methods and their performances are compared on the dataset. The best predictions are achieved by tree-based methods such as random forest and xgboost, and tuning the parameters of the algorithm helps to improve the classification accuracy. It is shown that the simultaneous use of data from cytotoxicity assays and from gene expression variables summarized in different ways has a synergistic effect and leads to a better prediction of hepatotoxicity than both sets of variables individually. Further, when gene expression data are summarized, different strategies for the generation of interpretable variables contribute to the overall improved prediction quality. When considering cytotoxicity assays alone, the best classification method yields a mean accuracy of 0.757, while the same classification method and an optimal choice of variables yields a mean accuracy of 0.811. The overall best value for the mean accuracy is 0.821.

This work estimates benchmarks for new approach method (NAM) performance in predicting organ-level effects in repeat dose studies of adult animals based on variability in replicate animal studies. Treatment-related effect values from the Toxicity Reference database (v2.1) for weight, gross, or histopathological changes in the adrenal gland, liver, kidney, spleen, stomach, and thyroid were used. Rates of chemical concordance among organ-level findings in replicate studies, defined by repeated chemical only, chemical and species, or chemical and study type, were calculated. Concordance was 39–88%, depending on organ, and was highest within species. Variance in treatment-related effect values, including lowest effect level (LEL) values and benchmark dose (BMD) values when available, was calculated by organ. Multilinear regression modeling, using study descriptors of organ-level effect values as covariates, was used to estimate total variance, mean square error (MSE), and root residual mean square error (RMSE). MSE values, interpreted as estimates of unexplained variance, suggest study descriptors accounted for 52–69% of total variance in organ-level LELs. RMSE ranged from 0.41 to 0.68 log₁₀-mg/kg/day. Differences between organ-level effects from chronic (CHR) and subchronic (SUB) dosing regimens were also quantified. Odds ratios indicated CHR organ effects were unlikely if the SUB study was negative. Mean differences of CHR - SUB organ-level LELs ranged from − 0.38 to − 0.19 log₁₀ mg/kg/day; the magnitudes of these mean differences were less than RMSE for replicate studies. Finally, in vitro to in vivo extrapolation (IVIVE) was employed to compare bioactive concentrations from in vitro NAMs for kidney and liver to LELs. The observed mean difference between LELs and mean IVIVE dose predictions approached 0.5 log₁₀-mg/kg/day, but differences by chemical ranged widely. Overall, variability in repeat dose organ-level effects suggests expectations for quantitative accuracy of NAM prediction of LELs should be at least ± 1 log₁₀-mg/kg/day, with qualitative accuracy not exceeding 70%.

The “Six Pack” is a set of animal toxicity studies that are widely used by industry and regulatory agencies to evaluate the toxicity of chemicals. It consists of three systemic toxicities (acute oral toxicity, acute inhalation toxicity and acute dermal toxicity) and three specific organ endpoints (eye damage/irritation, skin corrosion/irritation and skin sensitization). In the last two decades there has been a growing effort in the scientific community, as well as in regulatory agencies, to reduce and replace animal tests through implementation of alternative approaches. Computational methods in combination with in vitro measurements are pursued actively as the integrative approach for accurate and reliable assessment of chemical toxicity. Here, we generated structural alerts and developed a set of ten classification models for all six-pack endpoints using different molecular descriptors and machine learning techniques. The coverage of active chemicals by structural alerts was in the range from 24 % for acute inhalation toxicity to 52 % for acute oral toxicity. To establish confidence in model predictions, we used two different approaches to estimate the applicability domain (AD). The first approach was based on similarity distance between the query chemical and chemicals in the training set. In the second approach, the AD was estimated based on distance to model. The prediction accuracy of models evaluated using the validation sets was in the range from 0.67 for acute inhalation toxicity to 0.78 for acute dermal toxicity. The evaluation of models for chemicals within the similarity-based AD showed similar accuracy compared with the whole validation set. On the other hand, improvement of model performance was observed by using the distance to model approach to estimate AD, e.g. when distance to model was set to 0.3 the accuracy of predictions ranged from 0.75 for acute inhalation toxicity to 0.86 for acute oral toxicity. The combination of structural alerts and classification models provide a rapid means to screen a list of compounds for six-pack toxicity and to prioritize chemicals for in vitro toxicity evaluation.