A deep-learning approach for identifying prospective chemical hazards.

With the aim of helping to set safe exposure limits for the general population, various techniques have been implemented to conduct risk assessments for chemicals and other environmental stressors; however, none of these tools facilitate the identification of completely new chemicals that are likely hazardous and elicit an adverse biological effect. Here, we detail a novel in silico, deep-learning framework that is designed to systematically generate structures for new chemical compounds that are predicted to be chemical hazards. To assess the utility of the framework, we applied the tool to four endpoints related to environmental toxicants and their impacts on human and animal health: (i) toxicity to honeybees, (ii) immunotoxicity, (iii) endocrine disruption via ER-α antagonism, and (iv) mutagenicity. In addition, we characterized the predicted potency of these compounds and examined their structural relationship to existing chemicals of concern. As part of the array of emerging new approach methodologies (NAMs), we anticipate that such a framework will be a significant asset to risk assessors and other environmental scientists when planning and forecasting. Though not in the scope of the present study, we expect that the methodology detailed here could also be useful in the de novo design of more environmentally-friendly industrial chemicals.

The in silico identification of novel broad-spectrum antidotes for poisoning by organophosphate anticholinesterases.

Owing to their potential to cause serious adverse health effects, significant efforts have been made to develop antidotes for organophosphate (OP) anticholinesterases, such as nerve agents. To be optimally effective, antidotes must not only reactivate inhibited target enzymes, but also have the ability to cross the blood-brain barrier (BBB). Progress has been made toward brain-penetrating acetylcholinesterase reactivators through the development of a new group of substituted phenoxyalkyl pyridinium oximes. To help in the selection and prioritization of compounds for future synthesis and testing within this class of chemicals, and to identify candidate broad-spectrum molecules, an in silico framework was developed to systematically generate structures and screen them for reactivation efficacy and BBB penetration potential.

Adapting physiologically-based pharmacokinetic models for machine learning applications.

Both machine learning and physiologically-based pharmacokinetic models are becoming essential components of the drug development process. Integrating the predictive capabilities of physiologically-based pharmacokinetic (PBPK) models within machine learning (ML) pipelines could offer significant benefits in improving the accuracy and scope of drug screening and evaluation procedures. Here, we describe the development and testing of a self-contained machine learning module capable of faithfully recapitulating summary pharmacokinetic (PK) parameters produced by a full PBPK model, given a set of input drug-specific and regimen-specific information. Because of its widespread use in characterizing the disposition of orally administered drugs, the PBPK model chosen to demonstrate the methodology was an open-source implementation of a state-of-the-art compartmental and transit model called OpenCAT. The model was tested for drug formulations spanning a large range of solubility and absorption characteristics, and was evaluated for concordance against predictions of OpenCAT and relevant experimental data. In general, the values predicted by the ML models were within 20% of those of the PBPK model across the range of drug and formulation properties. However, summary PK parameter predictions from both the ML model and full PBPK model were occasionally poor with respect to those derived from experiments, suggesting deficiencies in the underlying PBPK model.

The in silico identification of novel broad-spectrum antidotes for poisoning by organophosphate anticholinesterases.

Because of their potential to cause serious adverse health effects, significant efforts have been made to develop antidotes for organophosphate (OP) anticholinesterases, such as nerve agents. To be optimally effective, antidotes must not only reactivate inhibited target enzymes, but also have the ability to cross the blood brain barrier (BBB). Progress has been made toward brain-penetrating acetylcholinesterase reactivators through the development of a new group of substituted phenoxyalkyl pyridinium oximes. To help in the selection and prioritization of compounds for future synthesis and testing within this class of chemicals, and to identify candidate broad-spectrum molecules, an in silico framework was developed to systematically generate structures and screen them for reactivation efficacy and BBB penetration potential.

kc-hits: a tool to aid in the evaluation and classification of chemical carcinogens.

MOTIVATION: The evaluation of chemicals for their carcinogenic hazard requires the analysis of a wide range of data and the characterization of these results relative to the key characteristics of carcinogens. The workflow used historically requires many manual steps that are labor-intensive and can introduce errors, bias and inconsistencies. RESULTS: The automation of parts of the evaluation workflow using the kc-hits software has led to significant improvements in process efficiency, as well as more consistent and comprehensive results. AVAILABILITY AND IMPLEMENTATION: https://gitlab.com/i1650/kc-hits.git. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

A Bayesian population physiologically based pharmacokinetic absorption modeling approach to support generic drug development: application to bupropion hydrochloride oral dosage forms.

We propose a Bayesian population modeling and virtual bioequivalence assessment approach to establishing dissolution specifications for oral dosage forms. A generalizable semi-physiologically based pharmacokinetic absorption model with six gut segments and liver, connected to a two-compartment model of systemic disposition for bupropion hydrochloride oral dosage forms was developed. Prior information on model parameters for gut physiology, bupropion physicochemical properties, and drug product properties were obtained from the literature. The release of bupropion hydrochloride from immediate-, sustained- and extended-release oral dosage forms was described by a Weibull function. In vitro dissolution data were used to assign priors to the in vivo release properties of the three bupropion formulations. We applied global sensitivity analysis to identify the influential parameters for plasma bupropion concentrations and calibrated them. To quantify inter- and intra-individual variability, plasma concentration profiles in healthy volunteers that received the three dosage forms, each at two doses, were used. The calibrated model was in good agreement with both in vitro dissolution and in vivo exposure data. Markov Chain Monte Carlo samples from the joint posterior parameter distribution were used to simulate virtual crossover clinical trials for each formulation with distinct drug dissolution profiles. For each trial, an allowable range of dissolution parameters ("safe space") in which bioequivalence can be anticipated was established. These findings can be used to assure consistent product performance throughout the drug product life-cycle and to support manufacturing changes. Our framework provides a comprehensive approach to support decision-making in drug product development.

Predicting the Disposition of the Antimalarial Drug Artesunate and Its Active Metabolite Dihydroartemisinin Using Physiologically Based Pharmacokinetic Modeling.

Artemisinin-based combination therapies (ACTs) have proven to be effective in helping to combat the global malaria epidemic. To optimally apply these drugs, information about their tissue-specific disposition is required, and one approach to predict these pharmacokinetic characteristics is physiologically based pharmacokinetic (PBPK) modeling. In this study, a whole-body PBPK model was developed to simulate the time-dependent tissue concentrations of artesunate (AS) and its active metabolite, dihydroartemisinin (DHA). The model was developed for both rats and humans and incorporated drug metabolism of the parent compound and major metabolite. Model calibration was conducted using data from the literature in a Bayesian framework, and model verification was assessed using separate sets of data. Results showed good agreement between model predictions and the validation data, demonstrating the capability of the model in predicting the blood, plasma, and tissue pharmacokinetics of AS and DHA. It is expected that such a tool will be useful in characterizing the disposition of these chemicals and ultimately improve dosing regimens by enabling a quantitative assessment of the tissue-specific drug levels critical in the evaluation of efficacy and toxicity.

Well-tempered MCMC simulations for population pharmacokinetic models.

A full Bayesian statistical treatment of complex pharmacokinetic or pharmacodynamic models, in particular in a population context, gives access to powerful inference, including on model structure. Markov Chain Monte Carlo (MCMC) samplers are typically used to estimate the joint posterior parameter distribution of interest. Among MCMC samplers, the simulated tempering algorithm (TMCMC) has a number of advantages: it can sample from sharp multi-modal posteriors; it provides insight into identifiability issues useful for model simplification; it can be used to compute accurate Bayes factors for model choice; the simulated Markov chains mix quickly and have assured convergence in certain conditions. The main challenge when implementing this approach is to find an adequate scale of auxiliary inverse temperatures (perks) and associated scaling constants. We solved that problem by adaptive stochastic optimization and describe our implementation of TMCMC sampling in the GNU MCSim software. Once a grid of perks is obtained, it is easy to perform posterior-tempered MCMC sampling or likelihood-tempered MCMC (thermodynamic integration, which bridges the joint prior and the posterior parameter distributions, with assured convergence of a single sampling chain). We compare TMCMC to other samplers and demonstrate its efficient sampling of multi-modal posteriors and calculation of Bayes factors in two stylized case-studies and two realistic population pharmacokinetic inference problems, one of them involving a large PBPK model.

pksensi: An R package to apply global sensitivity analysis in physiologically based kinetic modeling.

Sensitivity analysis (SA) is an essential tool for modelers to understand the influence of model parameters on model outputs. It is also increasingly used in developing and assessing physiologically based kinetic (PBK) models. For instance, several studies have applied global SA to reduce the computational burden in the Bayesian Markov chain Monte Carlo-based calibration process PBK models. Although several SA algorithms and software packages are available, no comprehensive software package exists that allows users to seamlessly solve differential equations in a PBK model, conduct and visualize SA results, and discriminate between the non-influential model parameters that can be fixed and those that need calibration. Therefore, we developed an R package, named pksensi, to make global SA more accessible in PBK modeling. This package can investigate both uncertainty and sensitivity in PBK models, including those with multivariate model outputs. It also includes functions to check the convergence of the global SA results. Overall, pksensi improves the user experience of performing global SA and can create robust and reproducible results for decision making in PBK model calibration.

Paraquat exposure and Parkinson's disease: A systematic review and meta-analysis.

To reconcile and unify available results regarding paraquat exposure and Parkinson's disease (PD), we conducted a systematic review and meta-analysis to provide a quantitative estimate of the risk of PD associated with paraquat exposure. Six scientific databases including PubMed, Cochrane libraries, EMBASE, Scopus, ISI Web of Knowledge, and TOXLINE were systematically searched. The overall odds ratios (ORs) with corresponding 95% CIs were calculated using a random-effects model. Of 7,309 articles identified, 13 case control studies with 3,231 patients and 4,901 controls were included into our analysis. Whereas, one prospective cohort studies was included into our systematic review. A subsequent meta-analysis showed an association between PD and paraquat exposure (odds ratio = 1.64 (95% CI: 1.27-2.13; I2 = 24.8%). There is a statistically significant association between paraquat exposure and PD. Thus, future studies regarding paraquat and Parkinson's disease are warranted.

Applying a Global Sensitivity Analysis Workflow to Improve the Computational Efficiencies in Physiologically-Based Pharmacokinetic Modeling.

Traditionally, the solution to reduce parameter dimensionality in a physiologically-based pharmacokinetic (PBPK) model is through expert judgment. However, this approach may lead to bias in parameter estimates and model predictions if important parameters are fixed at uncertain or inappropriate values. The purpose of this study was to explore the application of global sensitivity analysis (GSA) to ascertain which parameters in the PBPK model are non-influential, and therefore can be assigned fixed values in Bayesian parameter estimation with minimal bias. We compared the elementary effect-based Morris method and three variance-based Sobol indices in their ability to distinguish "influential" parameters to be estimated and "non-influential" parameters to be fixed. We illustrated this approach using a published human PBPK model for acetaminophen (APAP) and its two primary metabolites APAP-glucuronide and APAP-sulfate. We first applied GSA to the original published model, comparing Bayesian model calibration results using all the 21 originally calibrated model parameters (OMP, determined by "expert judgment"-based approach) vs. the subset of original influential parameters (OIP, determined by GSA from the OMP). We then applied GSA to all the PBPK parameters, including those fixed in the published model, comparing the model calibration results using this full set of 58 model parameters (FMP) vs. the full set influential parameters (FIP, determined by GSA from FMP). We also examined the impact of different cut-off points to distinguish the influential and non-influential parameters. We found that Sobol indices calculated by eFAST provided the best combination of reliability (consistency with other variance-based methods) and efficiency (lowest computational cost to achieve convergence) in identifying influential parameters. We identified several originally calibrated parameters that were not influential, and could be fixed to improve computational efficiency without discernable changes in prediction accuracy or precision. We further found six previously fixed parameters that were actually influential to the model predictions. Adding these additional influential parameters improved the model performance beyond that of the original publication while maintaining similar computational efficiency. We conclude that GSA provides an objective, transparent, and reproducible approach to improve the performance and computational efficiency of PBPK models.

A Novel Method for the Development of Environmental Public Health Indicators and Benchmark Dose Estimation Using a Health-Based End Point for Chlorpyrifos.

BACKGROUND: Organophosphorus (OP) compounds are the most widely used group of insecticides in the world. Risk assessments for these chemicals have focused primarily on 10% inhibition of acetylcholinesterase in the brain as the critical metric of effect. Aside from cholinergic effects resulting from acute exposure, many studies suggest a linkage between cognitive deficits and long-term OP exposure. OBJECTIVE: In this proof-of-concept study, we focused on one of the most widely used OP insecticides in the world, chlorpyrifos (CPF), and utilized an existing physiologically based pharmacokinetic (PBPK) model and a novel pharmacodynamic (PD) dose-response model to develop a point of departure benchmark dose estimate for cognitive deficits following long-term, low-dose exposure to this chemical in rodents. METHODS: Utilizing a validated PBPK/PD model for CPF, we generated a database of predicted biomarkers of exposure and internal dose metrics in both rat and human. Using simulated peak brain CPF concentrations, we developed a dose-response model to predict CPF-induced spatial memory deficits and correlated these changes to relevant biomarkers of exposure to derive a benchmark dose specific to neurobehavioral changes. We extended these cognitive deficit predictions to humans and simulated corresponding exposures using a model parameterized for humans. RESULTS: Results from this study indicate that the human-equivalent benchmark dose (BMD) based on a 15% cognitive deficit as an end point is lower than that using the present threshold for 10% brain AChE inhibition. This predicted human-equivalent subchronic BMD threshold compares to occupational exposure levels determined from biomarkers of exposure and corresponds to similar exposure conditions where deficits in cognition are observed. CONCLUSIONS: Quantitative PD models based on neurobehavioral testing in animals offer an important addition to the methodologies used for establishing useful environmental public health indicators and BMDs, and predictions from such models could help inform the human health risk assessment for chlorpyrifos. https://doi.org/10.1289/EHP1743.

Trichloroethylene-induced alterations in DNA methylation were enriched in polycomb protein binding sites in effector/memory CD4+ T cells.

Exposure to industrial solvent and water pollutant trichloroethylene (TCE) can promote autoimmunity, and expand effector/memory (CD62L) CD4+ T cells. In order to better understand etiology reduced representation bisulfite sequencing was used to study how a 40-week exposure to TCE in drinking water altered methylation of ∼337 770 CpG sites across the entire genome of effector/memory CD4+ T cells from MRL+/+ mice. Regardless of TCE exposure, 62% of CpG sites in autosomal chromosomes were hypomethylated (0-15% methylation), and 25% were hypermethylated (85-100% methylation). In contrast, only 6% of the CpGs on the X chromosome were hypomethylated, and 51% had mid-range methylation levels. In terms of TCE impact, TCE altered (≥ 10%) the methylation of 233 CpG sites in effector/memory CD4+ T cells. Approximately 31.7% of these differentially methylated sites occurred in regions known to bind one or more Polycomb group (PcG) proteins, namely Ezh2, Suz12, Mtf2 or Jarid2. In comparison, only 23.3% of CpG sites not differentially methylated by TCE were found in PcG protein binding regions. Transcriptomics revealed that TCE altered the expression of ∼560 genes in the same effector/memory CD4+ T cells. At least 80% of the immune genes altered by TCE had binding sites for PcG proteins flanking their transcription start site, or were regulated by other transcription factors that were in turn ordered by PcG proteins at their own transcription start site. Thus, PcG proteins, and the differential methylation of their binding sites, may represent a new mechanism by which TCE could alter the function of effector/memory CD4+ T cells.

Development of a physiologically based pharmacokinetic model of paraquat.

Paraquat (N, N'-dimethyl-4,4'-bipyridium dichloride) is a potent and widely used herbicide in agricultural countries, including Thailand. The presence of this chemical in the body can lead to toxic effects in the liver, kidney, and lung. Pulmonary toxicity has been identified as the main cause of acute toxicity in animals and humans. Chronic exposure to paraquat is associated with Parkinson's disease in humans. Paraquat is transported into the lungs by neutral amino acid transporter. Therefore, a physiologically based pharmacokinetic (PBPK) model of paraquat was developed with a description of the protein transporter mechanism. To develop a PBPK model of paraquat, a pharmacokinetic study of paraquat in rats was selected from the ThaiLIS and Pubmed database. The selected study contained tissue-specific concentration-time course information such as paraquat concentration levels in liver, kidney and lung. Physiologic parameters were acquired from the literature or determined using a Markov-Chain Monte Carlo (MCMC) technique. The developed PBPK model consisted of 5 organ compartments (i.e. kidney, liver, slowly perfused organs, richly perfuse organs and lung), featuring an incorporation of neutral amino acid transporter in the lung. Our model simulations could explain the data from the literature and adequately describe pharmacokinetics of paraquat in the rats. This developed PBPK model may be able help in understanding of paraquat-induced Parkinson's disease as well as in risk assessment of paraquat.

Characterizing the Effects of Race/Ethnicity on Acetaminophen Pharmacokinetics Using Physiologically Based Pharmacokinetic Modeling.

BACKGROUND AND OBJECTIVES: Acetaminophen (APAP, paracetamol) is currently the principal cause of acute liver failure in both the USA and the UK. However, relatively little is known about the influence of genes and race/ethnicity on the disposition of APAP and the extent to which genetic variation and ethnicity may predispose individuals to a higher risk of APAP-induced hepatotoxicity. The objective of this research was to develop subpopulation-specific physiologically based pharmacokinetic (PBPK) models for two genetically different groups (Western Europeans and East Asians) and then use the models to quantify the difference in absorption, distribution, metabolism, and excretion (ADME) of APAP between these groups. METHODS: A comprehensive set of human pharmacokinetic data mined from the literature was divided into two groups based on ethnicity as an indicator of the expected abundance of phenol-metabolizing alleles. Using these datasets and a Bayesian hierarchical framework, subpopulation-specific physiologically based pharmacokinetic models for APAP were developed and tested for the two groups. RESULTS: Model simulations were in good agreement with experimental data for both time-dependent parent and metabolite concentrations and summary pharmacokinetic parameters. In addition, simulations were conducted to characterize the difference between ADME in these groups with regard to urinary excretion and APAP area under the curve (AUC) in the liver. Although not dramatic at therapeutic dosing levels, these results demonstrated the divergence in the liver-specific APAP concentrations and AUC between the two groups and suggested that differences in glucuronidation capacity may play a role in this disparity. CONCLUSIONS: Overall, the models developed in this study, and others created using this type of hierarchical methodology, are expected to be useful in quantifying ADME in a subpopulation-specific manner and reducing prediction uncertainty compared to that from generalized PBPK modeling approaches.

Physiologically Based Pharmacokinetic Model of Rifapentine and 25-Desacetyl Rifapentine Disposition in Humans.

Rifapentine (RPT) is a rifamycin antimycobacterial and, as part of a combination therapy, is indicated for the treatment of pulmonary tuberculosis (TB) caused by Mycobacterium tuberculosis Although the results from a number of studies indicate that rifapentine has the potential to shorten treatment duration and enhance completion rates compared to other rifamycin agents utilized in antituberculosis drug regimens (i.e., regimens 1 to 4), its optimal dose and exposure in humans are unknown. To help inform such an optimization, a physiologically based pharmacokinetic (PBPK) model was developed to predict time course, tissue-specific concentrations of RPT and its active metabolite, 25-desacetyl rifapentine (dRPT), in humans after specified administration schedules for RPT. Starting with the development and verification of a PBPK model for rats, the model was extrapolated and then tested using human pharmacokinetic data. Testing and verification of the models included comparisons of predictions to experimental data in several rat tissues and time course RPT and dRPT plasma concentrations in humans from several single- and repeated-dosing studies. Finally, the model was used to predict RPT concentrations in the lung during the intensive and continuation phases of a current recommended TB treatment regimen. Based on these results, it is anticipated that the PBPK model developed in this study will be useful in evaluating dosing regimens for RPT and for characterizing tissue-level doses that could be predictors of problems related to efficacy or safety.

Chronic exposure to water pollutant trichloroethylene increased epigenetic drift in CD4(+) T cells.

AIM: Autoimmune disease and CD4(+) T-cell alterations are induced in mice exposed to the water pollutant trichloroethylene (TCE). We examined here whether TCE altered gene-specific DNA methylation in CD4(+) T cells as a possible mechanism of immunotoxicity. MATERIALS & METHODS: Naive and effector/memory CD4(+) T cells from mice exposed to TCE (0.5 mg/ml in drinking water) for 40 weeks were examined by bisulfite next-generation DNA sequencing. RESULTS: A probabilistic model calculated from multiple genes showed that TCE decreased methylation control in CD4(+) T cells. Data from individual genes fitted to a quadratic regression model showed that TCE increased gene-specific methylation variance in both CD4 subsets. CONCLUSION: TCE increased epigenetic drift of specific CpG sites in CD4(+) T cells.

A novel approach for estimating ingested dose associated with paracetamol overdose.

AIM: In cases of paracetamol (acetaminophen, APAP) overdose, an accurate estimate of tissue-specific paracetamol pharmacokinetics (PK) and ingested dose can offer health care providers important information for the individualized treatment and follow-up of affected patients. Here a novel methodology is presented to make such estimates using a standard serum paracetamol measurement and a computational framework. METHODS: The core component of the computational framework was a physiologically-based pharmacokinetic (PBPK) model developed and evaluated using an extensive set of human PK data. Bayesian inference was used for parameter and dose estimation, allowing the incorporation of inter-study variability, and facilitating the calculation of uncertainty in model outputs. RESULTS: Simulations of paracetamol time course concentrations in the blood were in close agreement with experimental data under a wide range of dosing conditions. Also, predictions of administered dose showed good agreement with a large collection of clinical and emergency setting PK data over a broad dose range. In addition to dose estimation, the platform was applied for the determination of optimal blood sampling times for dose reconstruction and quantitation of the potential role of paracetamol conjugate measurement on dose estimation. CONCLUSIONS: Current therapies for paracetamol overdose rely on a generic methodology involving the use of a clinical nomogram. By using the computational framework developed in this study, serum sample data, and the individual patient's anthropometric and physiological information, personalized serum and liver pharmacokinetic profiles and dose estimate could be generated to help inform an individualized overdose treatment and follow-up plan.

Physiologically based modeling of the pharmacokinetics of acetaminophen and its major metabolites in humans using a Bayesian population approach.

The principal aim of this study was to develop, validate, and demonstrate a physiologically based pharmacokinetic (PBPK) model to predict and characterize the absorption, distribution, metabolism, and excretion of acetaminophen (APAP) in humans. A PBPK model was created that included pharmacologically and toxicologically relevant tissue compartments and incorporated mechanistic descriptions of the absorption and metabolism of APAP, such as gastric emptying time, cofactor kinetics, and transporter-mediated movement of conjugated metabolites in the liver. Through the use of a hierarchical Bayesian framework, unknown model parameters were estimated using a large training set of data from human pharmacokinetic studies, resulting in parameter distributions that account for data uncertainty and inter-study variability. Predictions from the model showed good agreement to a diverse test set of data across several measures, including plasma concentrations over time, renal clearance, APAP absorption, and pharmacokinetic and exposure metrics. The utility of the model was then demonstrated through predictions of cofactor depletion, dose response of several pharmacokinetic endpoints, and the relationship between APAP biomarker levels in the plasma and those in the liver. The model addressed several limitations in previous PBPK models for APAP, and it is anticipated that it will be useful in predicting the pharmacokinetics of APAP in a number of contexts, such as extrapolating across doses, estimating internal concentrations, quantifying population variability, assessing possible impacts of drug coadministration, and, when coupled with a suitable pharmacodynamic model, predicting toxicity.