Virtual Screening Workflow

The Virtual Screening Workflow is designed to run an entire sequence of jobs for screening large collections of compounds against one or more targets. The workflow includes ligand preparation using LigPrep, filtering using ligfilter on QikProp properties or other structural properties, and Glide docking at the three accuracy levels, HTVS, SP, and XP. The design allows you to choose which of the stages to include in any run, and which selection of results from one stage are passed on to the next.

Before you run the workflow, you must ensure that receptors are properly prepared. See the Protein Preparation Guide — Contents for information on protein preparation. You can generate grids as part of the workflow, or you can generate them beforehand with Glide. See Receptor Grid Generation for information on grid generation.

The ligand files for the workflow can consist of 2D structures or 3D structures. If you supply 2D structures, you must run the ligand preparation part of the workflow to convert the structures to 3D for docking. This stage runs a LigPrep job, with the standard options for most parts of the LigPrep process. You can control some of these options, but if you want to use other options, you should run LigPrep on the ligands independently. For more information, see the LigPrep User Manual — Contents.

Prefiltering of the ligands can also be performed as part of the workflow. If you want to use the Lipinski Rule option, you must run QikProp to obtain the required properties. You can run QikProp as part of the workflow, regardless of the ligand source. If you have already run QikProp, you do not need to run it again. If you want to specify a custom filter, you can provide an input file for ligfilter, which will then filter the structures. For more information on ligfilter, see Filtering Structures by Property: ligfilter.

The full workflow includes three docking stages. The first stage performs HTVS docking. The ligands that are retained are then passed to the next stage, which performs SP docking; the survivors of this stage are passed on to the third stage, which performs XP docking. At each stage you can decide how many ligands are kept, and whether all the ionization and tautomeric states of each ligand are kept, or only some.

The workflow is intended to be robust, so that if a subjob fails, the master job will attempt to rerun the job a few times before quitting. If the master job fails (for example, due to a system failure or network failure) you can restart the job, and it will pick up the calculation from the latest point for which it has results and can restart from them.

If you publish work that uses VSW, please use the citation shown on our citation page.

The process of setting up a VSW job from Maestro is described in the remainder of this manual. If you want to run VSW jobs from the command line, see Running VSW Jobs from the Command Line.