I want to dock protein A to protein B using HADDOCK. I have a series of long distance (~20-30A) restraints generated from cross-linking mass spec (CXMS), and I think I can generate tbl file to be used in HADDOCK.
At the same time I also know several key residues in the interface of protein A and protein B from mutagenesis experiments, but I don’t know how to use the mutagenesis information in HADDOCK. Is that OK just to make the residues as “active residues” of the interface? Because I do not have chemical shift data, it’s impossible for me to figure out all of the residues in the interface. If I just make a few residues (one or two residues each protein) as “active residues” from the mutagenesis information, the ambiguous restraints generated by HADDOCK automatically from them would be too strict, according to my understanding (because no all of the interface residues are given to HADDOCK). But the mutagenesis information is useful to make sure the orientations of both proteins. How can I use the mutagenesis in formation in HADDOCK?
Yes you can translate your crosslinks into distance restraints. As for the mutagenesis data, you can indeed input them as active, but if you only have a few of those, you will need to define a large enough passive interface.
You might also give first a try to our new DISVIS webportal, which will allow you to visualize the interaction space define by your crosslinks. From it you could extract information about the possible interaction surfaces, which you can then input as passive to HADDOCK. For DISVIS visit:
In fact, I don’t really understand so called “passive residue” and “active residue”. According your reply, is the “passive residue” also located in the interface and “passive residues” include all of the “active residues”?
In the HADDOCK concept, an active residues should be part of the interface (will be forced to if possible)
A passive residue instead can be at the interface, but will not be penalised if not.
Better to be too generous in the definition of passive residues rather than too restrictive.
The DISVIS can output mrc density map representing the center of mass of the scanning chain conforming to the maximum found consistent restraints at every position in space. How to extract the possible surface from it? Just open the fixed chain an mrc map in molecular viewer and select the residues in the surface by hand?
Is there any method to generate a large passive interface automatically? Such as generating all of the residues accessible by solvent at a time?
And if I input a lot of passive residues (such as all of the residues in the surface) to generate AIR, will the calculation speed of HADDOCK be much slower?