Knowledge Base

Can I dock ligands to the whole protein, without specifying a particular binding site?

It is not possible to dock ligands to the entire protein surface ("blind" docking) in a single docking job. Glide docks ligands to a known or putative active site of the receptor. A spatially-focused search is necessary for efficiency, and the grid-based docking allows for a high-speed but extensive search of the active site. The region explored by the ligand center (the "inner box") is restricted to a cube whose sides cannot be longer than 40Å. However, if the cube is much larger than the default size of 10Å the chances of finding good poses actually decreases. The sides of the outer box, on which the grid is calculated, cannot be larger than 80Å, and the larger this box, the longer the docking job will take.

Instead, we recommend running SiteMap to identify possible binding locations. Grids can then be generated at those putative binding sites, and ligands can be docked to each of them. Our Virtual Screening Workflow can perform ensemble docking to multiple grids, so this is one way to dock ligands to multiple receptor locations in a single job. Alternatively, Glide Cross Docking can be used to run a SiteMap, Grid Generation, and Ligand Docking workflow.

The Glide Quick Start Guide includes a tutorial for generating a grid from SiteMap results (Docking to a Site Found by SiteMap).

If docking ligands to the entire protein surface is necessary, without limiting to the possible binding sites found by SiteMap, multiple grids would have to be generated with the grid centers chosen such that the grids cover the entire surface. Note that it is the inner grid boxes rather than outer grid boxes that should be tiled over the receptor surface. Unfortunately, we don't have any helper scripts to facilitate this, and it might be tedious to set up manually.

To ask a question or get help, please submit a support ticket or email us at

Back To Top