Siddharth Iyer (WITTRUP & CHURCH LABS)

Exploring the Virome for In Vivo RNA Delivery

Delivering gene editing RNA drugs to specific organs is difficult because RNA is unstable and most drug delivery vehicles encapsulating RNA go to the liver. We propose to look for viruses that naturally avoid the liver and go to specific organs and hijack these viruses as new ways to deliver RNA. We are developing new tools to design, create, interrogate, and iterate upon these delivery vectors. These tools will hopefully help enable the future of genetic medicine.

Emily Nieves (LAUFFENBURGER, EDELMAN & RACKAUCKAS LABS)

Uncertainty Quantified Discovery of Chemical Reaction Systems via Scientific Machine Learning

In this work we demonstrate an example of Bayesian scientific machine learning which aims to address the lack of interpretability and uncertainty quantification that traditional machine learning suffers. We extend a previously published scientific machine learning model that learns a reaction network from time course concentration data, the Chemical Reaction Neural Network (CRNN) to include a Bayesian framework. We demonstrate the Bayesian CRNN’s ability to recover the reaction network, and quantify the uncertainty in learned reaction rates.