Increased computational power provided by computer hardware has inspired algorithm developers to greatly expand the scope of tools for drug design. Today, computational chemists regularly make use of programs to model protein/ligand docking, search databases for small-molecule shape similarity, perform de novo protein design, build quantitative structure-activity relationship (QSAR) models, and calculate binding affinities. A cloud-computing platform can provide computational chemists with an ideal environment in which to deploy these tools. In this talk, we describe the deployment of a large-scale binding-affinity calculation tool called uf-BAC to the DNAnexus cloud-platform. uf-BAC performs multiple, independent molecular dynamics (MD) runs to create an ensemble of trajectories, from which accurate binding affinities can be deduced. By leveraging the elasticity offered by the cloud, users can easily launch one or many hundreds of simulations, and have the computational resources scale appropriately. Moreover, by providing a simplified web-frontend, users of various skill levels are able to build, monitor, or analyze uf-BAC results. Data sharing features allow the assignment of roles, granting collaborators appropriate access to view and analyze the data. Data is kept secure by strict enforcement of policies such as encrypting data at rest and in-flight. We believe that our experience in deploying uf-BAC to the cloud can serve as a model for others looking to migrate their tools and shows that the cloud will play an integral role in the future of computational chemistry.

Deploying large-scale binding-affinity calculations to the cloud

253rd National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.