The development of new strategies to treat neurodegenerative diseases is among the most challenging and expensive tasks for pharma. Various types of nanoparticles (NPs) are considered as versatile drug delivery systems to the brain. The “Fourth Industrial Revolution” brought novel digital solutions for drug discovery based on artificial intelligence (AI), including machine learning (ML). Unfortunately, the application of in silico methods in nanomedicine remains uncommon.

This presentation is aimed at highlighting the most recent achievements of NanoCARRIERS project aimed at integrating ML and physics-based molecular modeling to support designing brain drug delivery nanocarriers. We will demonstrate, how to use a combination of physics-based modeling (Density Functional Theory at the level of B3LYP) and ML for increasing the efficiency of predicting the energy of interaction between the surface of gold nanoparticles with proteins that can be used as targeting ligands. Moreover, will show a model that estimates stability of the considered structures in a physiological condition. Finally, we will report Quantitative Structure-Activity Relationships models for screening nanoparticles and ligands, based on their cytotoxicity and oxidative stress generation.

In addition to that, we will draw a perspective for further development of the field. The most important challenges related to the use of advanced deep learning and fundamental AI models will be discussed. Special attention will be put on the availability of training data and regulatory expectation for such models to be wider accepted as stand-alone or parts of New Approach Methodologies (NAMs).