ICCVAM Logo

ICCVAM Biennial Report 2016-2017

ICCVAM Biennial Report 2016-2017
Menu
https://ntp.niehs.nih.gov/go/849222

Computational Toxicology

Computational toxicology uses mathematics, informatics, and computer models to better understand toxicity mechanisms and predict toxic effects. ICCVAM agencies are exploring how these approaches could reduce and replace animal use for chemical safety testing.

DOD: Computational Rapid Identification Scientific Threat Analysis

The goal of the Defense Threat Reduction Agency’s Computational Rapid Identification Scientific Threat AnaLysis (CRISTAL) effort is to develop a computational approach to predict key attributes (such as physicochemical properties, environmental fate, and toxicity) of emergent threat agents. This approach will use a three-phase process to evaluate chemical properties, enabling a more rapid understanding of the relative threat of chemical substances than is possible using traditional laboratory testing. In the first phase, quantitative models will be used to predict physicochemical properties, which are then confirmed in limited laboratory evaluations. Second, chemical compounds with physicochemical properties that are consistent with a potential use as a threat agent will advance to the next stage, using algorithms to estimate a chemical’s environmental behavior, as well as its absorption, disposition, metabolism, and excretion profile. Third, computational estimates can then be confirmed and refined through laboratory experiments and limited animal validation studies.

NIEHS: Integrated Chemical Environment

Successful computational toxicology projects depend on high-quality data that are freely available and formatted for use in computational workflows. The NICEATM Integrated Chemical Environment (ICE) was developed to address the needs frequently expressed by NICEATM stakeholders.

As of the end of 2017, ICE included data from animal and non-animal tests that assessed regulatory endpoints such as acute oral toxicity, skin and eye irritation, skin sensitization, and endocrine activity. ICE also includes curated high-throughput screening data from Tox21 and physicochemical property data on chemicals, including solubility, melting point, and molecular weight. Data on EPA formulations allow users to compare labeling categories from EPA six-pack studies with the performance of the formulation’s active ingredients in non-animal methods. Downloadable workflows enable predictions of physicochemical properties, skin sensitization potency, and adverse outcome pathway mapping. ICE is open to all users with no registration needed.

Updates to ICE in 2018 will expand the in silico prediction models and computational workflows offered to include in vitro to in vivo extrapolation and characterization of chemicals from the ICE website, as well as through downloadable workflows. Other planned updates include the implementation of application programming interfaces to support availability and use of ICE data in computing environments outside of ICE. As part of the NIEHS mission to make data readily available, data in ICE are accessible through the NTP Chemical Effects in Biological Systems (CEBS) database, which will support use of ICE data in combination with other NIEHS data resources. Work is also ongoing to improve availability of ICE data to all NIH data resources.

NIEHS/EPA: Computational Tools for Physicochemical Properties Prediction

Physicochemical properties of chemicals are used as inputs for computational models that can predict a chemical’s potential toxicity, environmental fate, and exposure potential. However, experimental physicochemical property data are not available for many chemicals. NICEATM and EPA scientists developed computational models to rapidly estimate six physicochemical properties: octanol-water partition coefficient, water solubility, boiling point, melting point, vapor pressure, and bioconcentration factor. The performance of the new models was shown to be generally superior to existing resources. The computational models were described in a 2017 publication and are available via the NICEATM Integrated Chemical Environment and the EPA Chemistry Dashboard.

NIEHS/EPA: Computational Models for In Vitro to In Vivo Extrapolation

One key issue encountered with high-throughput in vitro testing methods is how to accurately relate chemical concentrations that induce in vitro responses to in vivo exposure concentrations that result in human or animal illness or injury. This relationship is established through in vitro to in vivo extrapolation (IVIVE), the topic of a 2016 workshop organized by NICEATM and the EPA National Center for Computational Toxicology. The workshop report was published in December 2017.

Computational toxicologists at NICEATM and collaborators in other branches of NTP and the National Center for Computational Toxicology are developing IVIVE analysis methods. Current work is focused on understanding the impact of various parameters, such as using free plasma concentration as a surrogate for the total plasma concentration and comparing multiple modeling approaches. Application of these IVIVE approaches to predict the potential of chemicals to cause developmental toxicity and interact with the endocrine system was described in presentations by Chang et al. at the 2017 Society of Toxicology annual meeting, 10th World Congress on Alternatives and Animal Use in the Life Sciences, and 2017 annual meeting of the American Society for Cellular and Computational Toxicology. A workflow for conducting these analyses is planned for a future release of the Integrated Chemical Environment (see article above).

Integrated Chemical Environment logo

Integrated Chemical Environment (ICE)

Launched in March 2017, the NICEATM Integrated Chemical Environment provides high-quality, curated data from NICEATM, its partners, and other resources, as well as tools to facilitate chemical safety assessment.