Skip to content.

Vakser Lab

Sections
Personal tools
You are here: Home » Our Resources » GRAMM » GRAMM-X Protein-Protein Docking Web Server » GRAMM-X Help Page

GRAMM-X Help

Document Actions

Interface residue constraints

Using this set of input fields, the user can restrict the docking to the approximately known location of the binding site. The user can describe:

  • which residues are likely to be present on the receptor-ligand interface
  • how rigorously the global set of predictions must be filtered in order to satisfy the interface residue constraints. 

In order to clarify the issue, we provide a representative example:

Suppose that the mutational studies or some other biochemical experiment showed that residues from 15 to 23 of the receptor (chain A) are critical for the interaction (the complex does not form if they are mutated), and the same is true for residues 86,87,112,115 of the ligand (chain B). So, there is a fair chance that those residues might be part of the receptor-ligand interface. At the same time, any of the following might take place: 

  • the list is not accurate:
    • some residues are listed simply because the biochemical method in question only provides an approximate sequence range
    • some of the listed residues are critical for the interaction, but not because they are in a surface contact with the partner protein
  • the list is not exhaustive - some interface residues are not known
  • the residues listed for the receptor might be in contact with some residues of the ligand that are not listed, and vice versa.

With this in mind, we could enter the following into the form:

Residue Constraints Options

Above, we requested GRAMM-X to restrict search to only those conformations that have: 

at least 5 of the listed receptor residues in contact with any residue of the ligand 

AND 

at least 2 of the listed ligand residues in contact with any residue of the receptor 

AND 

at least one receptor-ligand contact pair where both receptor and ligand residues come from the lists above. 

These numbers must reflect the quality of the biochemical information represented by the residue ranges. If a required number is too large and the residue range is too broad, the constraint might become too strict. That will cause all putative complex structures to be excluded. If the number of required residues is too small, the constraint will be weak and lead to more false positives in the output. The user might need to iteratively submit several docking jobs  in order to test stronger and weaker constraints while checking the output for the reasonably looking predictions.

Two residues are considered to be in contact if they have at least one pair of atoms within 4 Angstroms distance from each other.

Below are some examples that describe the format for the residue selection (quotation marks are used only to separate residue specifications from the text here and should not be entered in the actual web form):

  • '1-23:A' - residues from 1 to 23 (inclusive) from chain A
  • '40:A' - residue 40 from chain A
  • '1-23:' or '1-23' - residues 1-23 from chain(s) with an empty chain Id
  • '1-23:*' - residues 1-23 from all chains
  • '1A-10:B' - residues from 1A to 10 from chain B (A is an insertion code). If there are residues 10,10A, 10B, etc in the PDB file, then the above pattern will collect only residue 10.
  • '1-23:A,40:A,1A-10:B' - a comma-separated combination of any number of patterns.

The residue selection specifications are applied separately to the receptor and ligand molecules, so it is OK for the receptor and the ligand to share identical chain IDs.

 

Powered by Plone

This site conforms to the following standards: