I have a structure and a predicted binding site, and millions of compounds to dock, but no known actives. Is there any way to predict the things that are going to be active and exclude things that are almost certainly going to be inactive?

Probably the best approach to this problem is to use the Virtual Screening Workflow. This workflow involves running Glide HTVS, Glide SP, and Glide XP successively on your database, passing the top fraction of compounds on to subsequent stages. This multi-stage approach makes very large (million+ compound) screens possible in a reasonable time (the full database is docked only with the fastest mode), while still running Glide XP to generate the final results. Glide XP has additional penalty terms to help weed out false positives.

Keywords: actives, inactives, Virtual Screening Workflow, Glide XP, Glide SP

Back to Search Results