Friday, April 26, 2019 at 10:45 am Real time virtual screening of very large databases: Is bigger really better? Greg Warren, Ph.D.,OpenEye
Until recently most virtual screening, whether structure- or ligand-based, looked at compound databases on the order of 10 to maybe 30 times larger than what can be routinely screened experimentally by HTS. For database sizes in the low tens of millions, computational chemists could screen compounds using standard computational methods with on-premise hardware. Recently chemical suppliers, in particular Enamine and WuXi, have increased the number of rapidly and reliably purchasable compounds have a factor of hundreds going from millions to hundreds of millions or even billions. To screen these very large numbers of compounds, particularly in 3-D, new software and hardware strategies must be employed. This presentation will address how to virtually screen billions of molecules and will also discuss why screening databases that are hundreds to thousands of times larger might be beneficial.