Virtual screening, the computational discovery of new chemical matter with relevant biology, is a central tool in modern drug discovery. The substrates for virtual screening, libraries of chemically instantiated molecules (such as PubChem and
ZINC) or libraries of ‘virtual’ compounds, have been growing at unprecedented rates, putting increasing stress on existing virtual screening methods to keep up.
This webinar will introduce two revolutionary tools in virtual screening: FastROCS for searching by similarity in molecular shape and Orion, an integrated, cloud-native computational chemistry platform. By harnessing the compute power of the GPU
FastROCS has massively accelerated the discovery of novel, biologically active compounds in very large databases (FastROCS can search a database containing over 1.4 billion molecules in 15 minutes). Orion provides both the scalable, elastic,
on-demand compute resources required for these calculations, and the interactive analysis tools for compound selection, thereby bringing these ultra-large virtual screening calculations within reach of a very wide range of molecular designers,
not just expert computational chemists with access to substantial compute resources. This democratision of molecular design through the cloud eliminates the need for extensive experience in IT and HPC management, freeing the user to concentrate
on the important goal – the discovery of new medicines.
This educational webinar will be helpful for those involved with:
- High speed shape similarity calculations to search databases of billions of molecules in minutes, through a browser.
- The costs and benefits of calculation on the GPU.
- Complementarity between ligand-based and structure-based virtual screening methods.
Paul Hawkins PhD
Head of Scientific Solutions
Paul did his undergraduate degrees at the University of Southampton (chemistry and biochemistry) and his Ph.D. at “the home of golf”, St. Andrews (medicinal chemistry of HIV-1 protease). He was then a NATO post-doctoral fellow at Pennsylvania
State University, studying enzymatic hydroxylation.
Paul was a bench chemist in biotech, working mostly on anti-bacterial and gene regulation projects, before moving into the computational area as an applications scientist. He has worked at OpenEye for over a decade, the last 5 as Head of Scientific
Solutions, where he is instrumental in setting OpenEye’s research priorities and provides much of OpenEye’s public-facing science.
Paul’s main areas of interest at OpenEye are conformer generation, shape-based methods in drug discovery, and the application of statistical methods for the reliable comparison and validation of computational methods.