Case Study: Oak Ridge National Laboratory accelerates COVID-19 drug discovery with Google for Education

Tennessee team screens 1B virtual molecules in 20 hours to accelerate drug discovery for COVID-19

Oak Ridge National Laboratory (ORNL) needed a much faster way to discover potential COVID-19 treatments as the pandemic emerged, since traditional drug discovery can take more than 10 years and cost over $1 billion. Working with Google for Education, ORNL used Google Cloud research credits alongside its Summit supercomputer to build a high-throughput virtual screening pipeline for identifying promising drug candidates.

Google for Education helped ORNL run molecular docking at massive scale using AutoDock-GPU and parallel workloads across Summit and Google Cloud GPUs. The team screened 1.3 billion compounds in less than a day—about 50 times faster than with existing AutoDock software—and sent an initial batch of 323 top-scoring compounds for experimental testing, with more batches in progress.

Oak Ridge National Laboratory

Jeremy Smith

Director of the Center for Molecular Biophysics