Predictive Model of Rat Reproductive Toxicity from ToxCast High Throughput Screening

National Center for Computational Toxicology, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711, USA.
Biology of Reproduction (Impact Factor: 3.32). 05/2011; 85(2):327-39. DOI: 10.1095/biolreprod.111.090977
Source: PubMed


The U.S. Environmental Protection Agency's ToxCast research program uses high throughput screening (HTS) for profiling bioactivity and predicting the toxicity of large numbers of chemicals. ToxCast Phase I tested 309 well-characterized chemicals in more than 500 assays for a wide range of molecular targets and cellular responses. Of the 309 environmental chemicals in Phase I, 256 were linked to high-quality rat multigeneration reproductive toxicity studies in the relational Toxicity Reference Database. Reproductive toxicants were defined here as having achieved a reproductive lowest-observed-adverse-effect level of less than 500 mg kg(-1) day(-1). Eight-six chemicals were identified as reproductive toxicants in the rat, and 68 of those had sufficient in vitro bioactivity to model. Each assay was assessed for univariate association with the identified reproductive toxicants. Significantly associated assays were linked to gene sets and used for the subsequent predictive modeling. Using linear discriminant analysis and fivefold cross-validation, a robust and stable predictive model was produced capable of identifying rodent reproductive toxicants with 77% ± 2% and 74% ± 5% (mean ± SEM) training and test cross-validation balanced accuracies, respectively. With a 21-chemical external validation set, the model was 76% accurate, further indicating the model's potential for prioritizing the many thousands of environmental chemicals with little to no hazard information. The biological features of the model include steroidal and nonsteroidal nuclear receptors, cytochrome P450 enzyme inhibition, G protein-coupled receptors, and cell signaling pathway readouts-mechanistic information suggesting additional targeted, integrated testing strategies and potential applications of in vitro HTS to risk assessment.

Download full-text


Available from: Robert Kavlock, Nov 30, 2015
  • Source
    • "The Tox21 10 K chemical library consists of ∼10,500 plated compound solutions, consisting of 8311 unique chemical substances, including pesticides, industrial chemicals, food-use additives and drugs (Huang et al., 2014). Acquired activity data can serve not only as in vitro signatures that could be used to predict in vivo toxicity endpoints (Martin et al., 2011; Sipes et al., 2011) and to prioritize chemicals for extensive toxicity testing (Judson et al., 2010), but also to provide the scientific community with training data sets for developing reliable in silico toxicity models (Sun et al., 2012). Also, many attempts toward development of new computational methods for high-throughput toxicity prediction have been made and many techniques and algorithms have been proposed (Deeb and Goodarzi, 2012; Bakhtyari et al., 2013; Cheng et al., 2013; Valerio, 2013; Low et al., 2014; Omer et al., 2014; Toropov et al., 2014; Rouquie et al., 2015). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Toxicity evaluation of newly synthesized or used compounds is one of the main challenges during product development in many areas of industry. For example, toxicity is the second reason - after lack of efficacy - for failure in preclinical and clinical studies of drug candidates. To avoid attrition at the late stage of the drug development process, the toxicity analyses are employed at the early stages of a discovery pipeline, along with activity and selectivity enhancing. Although many assays for screening in vitro toxicity are available, their massive application is not always time and cost effective. Thus the need for fast and reliable in silico tools, which can be used not only for toxicity prediction of existing compounds, but also for prioritization of compounds planned for synthesis or acquisition. Here I present the benchmark results of the combination of various attribute selection methods and machine learning algorithms and their application to the data sets of the Tox21 Data Challenge. The best performing method: Best First for attribute selection with the Rotation Forest/ADTree classifier offers good accuracy for most tested cases. For 11 out of 12 targets, the AUROC value for the final evaluation set was ≥0.72, while for three targets the AUROC value was ≥ 0.80, with the average AUROC being 0.784±0.069. The use of two-dimensional descriptors sets enables fast screening and compound prioritization even for a very large database. Open source tools used in this project make the presented approach widely available and encourage the community to further improve the presented scheme.
    Preview · Article · Dec 2015
  • Source
    • "This includes the development of in vitro testing databases for hundreds of assays and thousands of chemicals in the ToxCast and Tox21 programs (Attene-Ramos et al. 2013; Kavlock et al. 2012). This has fostered development of computational models predicative of in vivo adverse outcomes (Martin et al. 2011; Rotroff et al. 2013; Sipes et al. 2011). The lack of adequate data sets for large numbers of chemicals from in vitro DNT assays has severely hampered the development of computational models. "
    [Show abstract] [Hide abstract]
    ABSTRACT: A major problem in developmental neurotoxicity (DNT) risk assessment is the lack of toxicological hazard information for most compounds. Therefore, new approaches are being considered to provide adequate experimental data that allow regulatory decisions. This process requires a matching of regulatory needs on the one hand and the opportunities provided by new test systems and methods on the other hand. Alignment of academically and industrially driven assay development with regulatory needs in the field of DNT is a core mission of the International STakeholder NETwork (ISTNET) in DNT testing. The first meeting of ISTNET was held in Zurich on 23-24 January 2014 in order to explore the concept of adverse outcome pathway (AOP) to practical DNT testing. AOPs were considered promising tools to promote test systems development according to regulatory needs. Moreover, the AOP concept was identified as an important guiding principle to assemble predictive integrated testing strategies (ITSs) for DNT. The recommendations on a road map towards AOP-based DNT testing is considered a stepwise approach, operating initially with incomplete AOPs for compound grouping, and focussing on key events of neurodevelopment. Next steps to be considered in follow-up activities are the use of case studies to further apply the AOP concept in regulatory DNT testing, making use of AOP intersections (common key events) for economic development of screening assays, and addressing the transition from qualitative descriptions to quantitative network modelling.
    Full-text · Article · Jan 2015 · Archive für Toxikologie
  • Source
    • "Reproductive and developmental toxicity were even estimated to become the largest animal user for safety testing within REACH (Pedersen et al., 2003; Van der Jagt et al., 2004) since approximately 10,000 chemicals with an annual volume of >100 tonnes would have to be tested on reproductive toxicity. The estimates ranged from 40% to 90% of the total number of animals to comply with REACH that would be needed for reproductive toxicity testing purposes (Van der Jagt et al., 2004; Spielmann and Vogel, 2006; Hartung and Rovida, 2009; Martin et al., 2011). At about the same time, several studies became available that questioned the added value of the second generation (Cooper et al., 2006; Janer et al., 2007a,b; Martin et al., 2009; Piersma et al., 2011) and criticized the limited predictive value of the OECD TG 416 for developmental immunotoxic and neurotoxic parameters (See Section 2.1.). "
    [Show abstract] [Hide abstract]
    ABSTRACT: The two-generation study (OECD TG 416) is the standard requirement within REACH to test reproductive toxicity effects of chemicals with production volumes >100 tonnes. This test is criticized in terms of scientific relevance and animal welfare. The Extended One Generation Reproductive Toxicity Study (EOGRTS), incorporated into the OECD test guidelines in 2011 (OECD TG 443) has the potential to replace TG 416, while using only one generation of rats and being more informative. However, its regulatory acceptance proved challenging. This article reconstructs the process of regulatory acceptance and use of the EOGRTS and describes drivers and barriers influencing the process. The findings derive from literature research and expert interviews. A distinction is made between three sub-stages; The stage of Formal Incorporation of the EOGRTS into OECD test guidelines was stimulated by retrospective analyses on the value of the second generation (F2), strong EOGRTS advocates, animal welfare concern and changing US and EU chemicals legislation; the stage of Actual Regulatory Acceptance within REACH was challenged by legal factors and ongoing scientific disputes, while the stage of Use by Industry is influenced by uncertainty of registrants about regulatory acceptance, high costs, the risk of false positives and the manageability of the EOGRTS.
    Full-text · Article · Oct 2014 · Regulatory Toxicology and Pharmacology
Show more