Authors
Bhati, Agastya; Wan, Shunzhou; Alfe, Dario; Clyde, Austin; Bode, Mathis; Tan, Li; Titov, Mikhail; Merzky, Andre; Turilli, Matteo; Jha, Shantenu; Ma, Heng; Trifan, Anda; Ramanathan, Arvind; Brettin, Tom; Partin, Alexander; Xia, Fangfang; Stevens, Rick
Abstract
The race to meet the challenges of the global pandemic has served as a reminder thatthe existing drug discovery process is expensive, inefficient and slow. There is a majorbottleneck screening the vast number of potential small molecules to shortlist leadcompounds for antiviral drug development. New opportunities to accelerate drug discovery lie at the interface between machine learning methods, in this case developedfor linear accelerators, and physics-based methods. The two in silico methods, eachhave their own advantages and limitations which, interestingly, complement each other.Here, we present an innovative infrastructural development that combines bothapproaches to accelerate drug discovery. The scale of the potential resulting workflowis such that it is dependent on supercomputing to achieve extremely high throughput.We have demonstrated the viability of this workflow for the study of inhibitors for fourCOVID-19 target proteins and our ability to perform the required large-scalecalculations to identify lead antiviral compounds through repurposing on a variety ofsupercomputers.