CORAL and Nano-QFAR: Quantitative feature - Activity relationships (QFAR) for bioavailability of nanoparticles (ZnO, CuO, Co3O4, and TiO2)

- The predictive model for cell viability is developed;
- The predictive value is calculated with quasi-SMILES;
- In contrast to traditional SMILES quasi-SMILES is representation of conditions;
- Each condition is represented by a code;
- Optimal descriptor is sum of correlation weights of the codes of conditions.

Quantitative feature - activity relationships (QFAR) approach was applied to prediction of bioavailability of metal oxide nanoparticles. ZnO, CuO, Co3O4, and TiO2 nanoxides were considered. The computational model for bioavailability of investigated species is asserted. The model was calculated using the Monte Carlo method. The CORAL free software (http://www.insilico.eu/coral) was used in this study. The developed model was tested by application of three different splits of data into the training and validation sets. So-called, quasi-SMILES are used to represent the conditions of action of metal oxide nanoparticles. A new paradigm of building up predictive models of endpoints related to nanomaterials is suggested. The paradigm is the following “An endpoint is a mathematical function of available eclectic data (conditions)”. Recently, the paradigm has been checked up with endpoints related to metal oxide nanoparticles, fullerenes, and multi-walled carbon-nanotubes.