Poster: nanoQSAR modelling

nanoQSAR modelling using protein corona fingerprints

Georgia Tsiliki, Haralambos Sarimveis / National Technical University of Athens


Georgia Tsiliki

OpenTox Euro 2014 Poster: nanoQSAR modelling using protein corona fingerprints

Recent studies have shown that the presence of serum proteins within in vitro cell culture systems forms a protein adsorption layer (a.k.a. the "protein corona") on the surface of nanoparticles that affects nanoparticle-cell interactions and cell response [1, 2].

The protein corona thus encodes information about the interface formed between the nanoparticle and the cell surface within a physiological environment. Here, we analyze a number of recently published serum protein corona ‘fingerprints’ formed around a library of 105 surface-modified gold nanoparticles [3].

The models presented aggregate relative abundances of spectral counts with gene ontology information, particularly the gene ontology information specific to each protein corona are used to calculate a new set of descriptors, referred to as GO descriptors.

Additionally, parameters extracted from nanoparticle characterization assays (e.g. nanoparticle size, aggregation state, and surface charge) are considered.

Our goal is to enrich the data using gene set information whlist emphasizing the importance of -omics data in modelling toxicity. Different QSAR models are compared in terms of coefficient of determination, Root mean squared error, and Akaike information criterion.

To improve the prediction accuracy of the model we employ bootstrapping techniques. The final set of GO descriptors estimated by nanoQSAR models is further exploited for their biological relevance and functional similarity.

1. Ge, C.; Du, J.; et al. Proc. Natl. Acad. Sci. U. S. A., 108, 2011, 16968-16973.
2. Lesniak, A.; Fenaroli, F.; et al. ACS Nano, 6, 2012, 5845-5857.
3. Walkey, C.D.; Olsenb, J.B.; et al. ACS Nano, 8 (3), 2014, 2439–2455.