In Silico Methods for the Derivation of Quantitative Safety Limits (S13-7)

In silico toxicology approaches can support chemical safety assessment by providing effective and convenient predictions of chemicals’ toxicity. They generally embrace different methodologies including statistical-based methods, expert systems (i.e., structural alerts) as well as grouping and read-across. They are generally referred to as (Q)SARs, i.e., (Quantitative) Structure Activity Relationships, and their application for the derivation of safety regulatory limits is a valuable resource meeting the demand for risk characterization of an ever-growing number of chemicals, while pursuing the 21st century toxicology vision.
In silico methods can be used to generate toxicity predictions that drive the categorization of the target chemical within a given class associated with a specific limit. On the other hand, (Q)SARs can also be used to support the derivation of quantitative regulatory limits like PDE (Permitted Daily Exposure) limits. In this case, in silico toxicity profiles can be generated for the source and the target substance in order to support a read-across justification; they may thus help to establish whether the target substance presents additional elements of concern with respect to the potential analogue. This analysis can then support the evaluation of whether the source is representative of the toxicity of the target substance and whether the toxicity data of the source can be used for reading-across the quantitative safety limit to the target substance.
The interplay of the different in silico toxicology approaches is critical for regulatory applications. Importantly, knowledge-base and continuous refinement of (Q)SARs are at the core of their advancement.