Decoding the Systems Pharmacology of Proteinopathy in Human Cells

Large-scale chemical screening against prion-like proteins in human cells is a powerful way to discover and understand therapeutic candidates for neurodegenerative diseases. However, the ways that new compounds work and the routes to advance them to the clinic are unclear. This is challenging because we must combine large cellular screen datasets, tangled webs of drug-target interactions, and a diverse array of possible neurodegenerative mechanisms into actionable information. Drug-target binding informs therapeutic compound development from the bottom up, and cellular screening does so from the top down: where they meet, each meaningfully constrains the solutions of the other. Consequently, we will determine the biological mechanisms of promising new drug-like compounds using interpretable deep learning, to improve understanding of proteinopathies and accelerate drug discovery for neurodegenerative diseases.