Machine Learning for Therapeutic Discovery
Sepsis, a manifestation of the body’s inflammatory response to injury and infection, has a mortality rate of between 28%-50% and affects approximately 1 million patients annually in the United States. Currently, there are no therapies targeting the cellular/molecular processes driving sepsis that have demonstrated the ability to control this disease process in the clinical setting. We propose that this is in great part due to the considerable heterogeneity of the clinical trajectories that constitute clinical “sepsis,” and that determining how this system can be controlled (through the use of dynamic, adaptive, and novel therapeutic strategies) back into a state of health requires the application of concepts drawn from the field of dynamical systems.
In order to discover successful potential therapies, we use techniques in Machine Learning to operate on our computational models, in effect, treating them as proxies for the biological system they represent, allowing us to overcome hurdles (cost, ethics, etc.) involved with live experimentation and data collection.
1. Cockrell, Robert Chase, and Gary An. "Examining the controllability of sepsis using genetic algorithms on an agent-based model of systemic inflammation." PLoS computational biology 14, no. 2 (2018): e1005876.
2. Petersen, Brenden K., Jiachen Yang, Will S. Grathwohl, Chase Cockrell, Claudio Santiago, Gary An, and Daniel M. Faissol. "Precision medicine as a control problem: Using simulation and deep reinforcement learning to discover adaptive, personalized multi-cytokine therapy for sepsis." arXiv preprint arXiv:1802.10440 (2018).