Review of (Q)SAR models for regulatory assessment of nanomaterials risks

A comprehensive understanding of the relationships between the physicochemical properties and the behaviour of nanomaterials in biological systems is required for designing safe and functional nanomaterials. Quantitative Structure-Activity Relationship (QSAR) methods help to establish such relationships, and their growing importance for providing key information is reflected in a number of regulatory frameworks (e.g., Registration Evaluation Authorization and Restriction of Chemical Substances (REACH)), where these approaches are considered acceptable, under certain conditions, for filling in knowledge gaps for untested chemicals. The (Q)SAR methodology is well known and extensively applied in the areas of drug discovery and chemical toxicity modelling for guiding the experimental design of chemical compounds, but its application to nanomaterials still requires major advancements. This review analyses the currently available (Q)SAR models for regulatory assessment of nanomaterials risks. The survey first analyzes and discusses the regulatory relevance and reliability of models within the context of the 5 OECD (Q)SAR validation criteria. Then, the information required to apply the models is outlined, focusing especially on descriptor development. The most important mechanisms of toxicity, knowledge gaps that impede the development of structure-activity models, as well as the domains where (Q)SAR models are expected to convey fundamental contributions, such as in tiered testing schemes and analysis of functional assays' results, are also discussed. It is concluded that the development of (Q)SAR models for nanomaterials requires both the development of nano-specific descriptors and curated experimental datasets.