EU Science Hub

Increasing confidence in computational safety assessment of chemicals

Computer code
Nov 22 2018

JRC scientists are part of an international initiative developing principles and protocols for the consistent use of computational models in chemical safety assessment to promote greater acceptance in regulatory applications.

The In Silico Toxicology (IST) Protocol initiative is focusing on how results from computational methods should be generated, interpreted, assessed and documented with a view to increasing confidence in their use.

In its first publication, the IST consortium describes the general framework of IST Protocols and how they can guide the integration of computational predictions with experimental data to support the assessment of a chemical for adverse health effects.

Computational (“in silico”) models provide valuable information on chemical and toxicological properties and are very useful in chemical safety assessment, for example in the context of drug development or the toxicity screening of industrial chemicals.

Commonly used in silico approaches include models that infer biological activity from features of a chemical's structure ("structure-activity relationships") or that extrapolate properties between similar substances ("read-across"). Other models take the 3D molecular shape into account, for example to predict docking of molecules to biological receptors. In a regulatory context, in silico methods are routinely applied in the assessment of drug impurities and are used by many registrants as an alternative to animal testing within the context of REACH. However, the regulatory acceptance of in silico approaches is still limited and one reason for this is the absence of harmonised protocols on how to properly apply methods and interpret results.

The IST Protocols include consideration of the reliability and relevance assessment of computational methods using a novel scoring scheme and the conduct and documentation of expert review using practical checklists.

Work is ongoing to apply the framework to specific toxicological endpoints, such as genotoxicity, skin sensitisation, and acute and repeated dose systemic toxicity. Ultimately, use of the framework and associated protocols is expected to increase the acceptance and uptake of in silico methodologies, thereby strengthening the protection of human health and the environment while reducing reliance on animal testing.

Read more in:

Myatt G.J. et al., “In silico toxicology protocols”, Regulatory Toxicology and Pharmacology, 96 (2018), pp. 1-17. doi: 10.1016/j.yrtph.2018.04.014

JRC wide hidden block