Robust Deep Image Embeddings for Drug Repurposing
At Recursion, we are developing one of the richest fluorescent microscopy image sets in the world using Deep Learning. Unlike many computer vision tasks ours is not one of classification or recognition, but rather is focused on learning abstract representations of thousands of experimental conditions in order to identify drugs as potential treatments for rare genetic diseases. In this presentation, I will demonstrate an approach that uses a combination of the ResNet architecture along with a triplet loss to learn robust abstract representations of our cell populations to accomplish this goal.
As Recursion Pharmaceutical's lead data scientist, Mason is focused on upending the pharmaceutical industry by using fluorescent microscopy imaging to scalably repurpose drugs to treat rare genetic diseases. Prior to joining Recursion, Mason earned his MS in Mathematics from BYU, where he focused on developing new approaches to identify individuals at-risk for chronic kidney disease. After an internship with the NSA, being unable to decide between becoming a spook or pursuing a PhD he chose the obvious route: join a start-up instead, where he built systems to mine unstructured medical records, analyze human behavior through cell phone logs, and simulate complex call centers.