In silico Toxicity Assessments ((Q)SAR)

In silico toxicity predictions, including (quantitative) structure-activity relationships ((Q)SAR) and read-across, can be useful tools at all stages of active substance development. During research and development (R&D) high-throughput screening using various in silico tools can allow the relative toxicity of a large series of substances to be rapidly evaluated. The least toxic substances can be prioritised for further development. In addition, substances with potential toxicological concern are identified early in development, informing the correct testing to be performed to evaluate the accuracy of the prediction.
In silico toxicity predictions can also be used as part of product lifecycle management for substances under review. (Q)SARs are required under some regulatory guidelines, including residue definition (EFSA guidance on the establishment of the residue definition for dietary risk assessment) and impurities (guidance document on the assessment of the equivalence of technical materials of substances regulated under regulation (EC) No 1107/2009). The main endpoint of interest for these guidelines is genotoxicity, for which very accurate predictions (>85% accuracy) can be made using a combination of (Q)SAR and read-across techniques.

Performing in silico predictions, at any stage of development, can provide fast results that are cheaper than traditional toxicity testing methods and do not require the use of animals. Furthermore, in silico predictions allow substances that cannot easily be synthesised/isolated (e.g., metabolites and impurities) to be evaluated.

At Exponent we have expertise in the use and interpretation of a variety of in silico toxicity tools. These include Derek Nexus, Leadscope, OECD QSAR Toolbox, TEST, VEGA, ToxTree, ECOSAR, and EPISUITE. In addition, we have complimentary expertise in genotoxicity that allows Exponent to provide detailed analysis of the implications of (Q)SAR outcomes and testing strategies to address them.

Professionals

Knowledge