Universiteit Leiden

nl en


Development of Quantitative Nanostructure Activity Relationship (QNAR) Models Predicting the Toxicity of Metal-based Nanoparticles to Aquatic Species

Describe and identify what dosimetry parameters are of importance to interpret dose-response relationships (eg., mortality, sub-lethal, growth or reproduction inhabitation, DNA damage and reactive oxygen species, etc. ) for metal-based nanoparticles?
How to develop quantitative models that enable to predict ecotoxicological effects based on selected nanoparticle properties?
Is there any possibility to extrapolate predictive models between species, species groups (species from similar taxonomic groups or similar traits) or over different biological organization levels?

2013  -   2017
Guangchao Chen
China Scholarship Council China Scholarship Council

The project links very well with the VIDI-project of dr. Martina G. Vijver (received in 2014) entitled: The added risk of size at the nano-bio interface: quantifying uptake, internal localization and partitioning of metal-based nanoparticles in aquatic organisms.

Short abstract

Given the almost exponential growth of nanoparticles, it is pivotal for unhindered industry-driven development of engineered nanomaterials (ENMs) that scientifically justified predictive models and modeling techniques become available allowing for accurate screening of the potential adverse effects. This work aims at the development and validation of models to predict the effect of ENMs in aquatic media. Ecotoxicological responses of the ENMs will be modeled based on a state-of-the-art characterization of ENMs, state-of-the-art modeling techniques, and development of database on ENMs ecotoxicity test results. The consequent validation will be performed by experimental testing using aquatic organisms.

Project description

Computational modeling has emerged over the past decade as a reliable tool to estimate the parameters that control properties and effects of chemical substances on the basis of (quantitative) structure-activity relationship (Q)SAR and read-across techniques. Combined with powerful data-mining tools, these computational models offer a rapid way of filling data gaps due to limited availability or even due to complete lack of experimental data on new substances. Only a few attempts have recently been made to develop preliminary models by relating nanoparticle properties to effects.

These preliminary attempts have not only indicated the tantalizing possibility that a QSAR approach may indeed be feasible and useful in predicting the physicochemical properties and effects of ENMs but also revealed the challenges facing QSAR modeling of nanomaterial toxicity and areas that need research. QSAR modeling of the fate and effects of ENMs is a novel area of research, especially given the global desire of minimizing animal testing and reducing costs of regulatory testing of chemicals. Therefore the PhD-research seeks to investigate the application of different (Q)SAR approaches to the novel research area of nanomaterial toxicity modeling.

The strategy of building QSAR models for ENMs include:

  1. Collection of experimental data based on publications and databases;
  2. Evaluation of retrieved data
  3. Model construction using different modeling techniques;
  4. Evaluation of constructed models
  5. Models validation on basis of experimental assays

Chen G., Li X., Chen J., Zhang Y. & Peijnenburg W.J.G.M. (2014), Comparative study of biodegradability prediction of chemicals using decision trees, functional trees, and logistic regression, Environmental Toxicology and Chemistry 33(12): 2688-2693.