Scientific Reports, v. 9, 15465, 2019.

Abstract: The development of nanomedicines currently suffers from a lack of efficient tools to predict pharmacokinetic behaviour without relying upon testing in large numbers of animals, impacting success rates and development costs. The aim of this work was therefore to present dendPoint, the first in silico model to predict the intravenous pharmacokinetic profiles of a commonly explored drug vector, based on dendrimer physicochemical properties.

In this instance, we compiled a relational database of key pharmacokinetic parameters and structural/physicochemical properties of dendrimers from literature reports used to develop a machine learning-based predictor capable of accurately predicting pharmacokinetic properties of dendrimer-based constructs including Half-life, Clearance, Volume of Distribution and Liver/Urine Doses. We show dendPoint successfully predicts pharmacokinetics properties achieving correlations of up to r = 0.83.

We believe and expect this platform can ultimately be used to guide dendrimer construct design and refinement prior to embarking on more lengthy and expensive efforts to develop and test in vivo behaviour.