A Stochastic Multiscale Modeling Platform for Targeted Drug Delivery

Arif Masud, University of Illinois at Urbana-Champaign

Usage Details

Targeted drug delivery using nano-sized carriers aims at achieving maximum efficacy with minimum dosage of medicine. However, this technology is plagued with inherent uncertainty and large dimensionality of the stochastic biological systems. We have developed advanced numerical methods that can reduce the stochastic dimensionality via novel fine-scale modeling concepts. We will apply it to optimize particle size and shape, and maximize the adhesion of drug carriers to a target vascular wall. The virtual module will help train the medical students at Carl-Illinois College of Medicine, where PI Masud has an affiliate appointment, to test various patient specific drug delivery strategies. From Blue Waters perspective, novel feature will be the exploitation of the local resident memory and hardware architecture at the level of processing nodes by increasing the intensity of local solves while reducing the size of global problem, and embedding of Machine Learning ideas at the subgrid level to help augment physics based models with data-driven models.