Quantitative Structure-Activity-Relationships for cellular uptake of nanoparticles

Abstract

Quantitative Structure-Activity-Relationships (QSARs) were investigated for cellular uptake of nanoparticles (NPs) using a dataset comprised of 109 NPs of the same iron oxide core but with different surface-modifying organic molecules. QSARs were built using both linear and non-linear model building methods along with a forward descriptor selection from an initial pool of 184 chemical descriptors calculated for the NP surface-modifying organic molecules. The resulting QSAR was a robust Relevance Vector Machine (RVM) model built with nine descriptors, which demonstrated prediction accuracy as quantified by a 5-fold cross-validated squared correlation coefficient (RCV2) of 0.77. The William's plot for the RVM based QSAR shows that the nine selected descriptors spanned a reasonable applicability domain. The developed QSAR can provide useful insight regarding parameters that affect NP cellular uptake and thus provide guidance for the selection and/or design of NPs for biomedical applications.