Connor Coley is reshaping how scientists hunt for new medicines. With estimates of 10^20 to 10^60 possible small-molecule compounds out there, testing them all the old way would take forever. That's why his team at MIT uses artificial intelligence to spot promising drug candidates fast.
From family lab roots to faculty
Science runs in Coley's blood, dad’s a radiologist, mom studied molecular biophysics, grandma taught math. He finished high school at 16, majored in chemical engineering at Caltech, then earned his PhD at MIT. His early work mixed machine learning with cheminformatics to plan reactions and automate synthesis under the DARPA Make-It program.
At just 25 he joined the MIT faculty, drawn by the place’s unique cross-department energy. After a postdoc at Broad Institute focused on DNA-encoded libraries, he launched his lab in 2020.
Smart Models That Think Like Chemists
Coley’s group builds tools that pair real chemistry challenges with clever computation.
ShEPhERD scores new molecules by how their 3D shapes interact with target proteins, now helping pharma companies find drugs.
FlowER, a generative model, predicts reaction products while respecting mass conservation and reaction mechanisms. “We’ve spent a lot of time thinking about how to make sure that our machine-learning models are grounded in an understanding of reaction mechanisms, in the same way an expert chemist would be.”
His students also tackle structure elucidation, lab automation, and experiment design.
Why AI is changing chemistry forever
This intersection of code and flasks is speeding up everything from drug design to new molecule creation. Coley sums it up simply: it’s a general approach for any organic molecule application, with small-molecule drugs at the top of the list.
