Repurposing High-Throughput Imaging Tests for Drug Discovery Allows for Biological Activity Prediction

A glucocorticoid receptor High-Throughout (cell) Imaging (HTI) screen to predict target protein activity in a variety of seemingly unrelated tests. The repurposing method raised hit rates by 60 to 250 times over the primary project assays in two ongoing drug discovery projects, while also boosting the chemical structure diversity of the hits using machine learning. Our findings show that data from existing HTI screens is a significant source of data that may be repurposed to aid drug discovery efforts.