Multidomain docking and antibody CDR loop flexibility?

HADDOCK
1

Hi!

I have the following questions regarding the usage of HADDOCK for wrapping up two of my ongoing projects. Any inputs would be highly appreciated.

  1. we want to model protein A-B complex, and both have a solved crystal structure respectively. Protein A structure is a tight conformation which is known to undergo drastic conformational change upon binding to its known partner. Now we have solid crosslinked restraint evidence connecting 3 out of 7 domains of Protein A with Protein B. Is there any suggestion regarding how to dock these 3 domain separately to Protein B and combine them together afterwards in 1:1 complex using HADDOCK? I have generated both DisVis-derived interface for those 3 domains-Protein B and corresponding distance restraints.

  2. I want to generate a antibody-antigen complex where antibody is modelled through AF2 multimer. DIsVis-derived interaction analysis provide a set of active residues. Do I need to specify which segment in CDRs to be semi-flexible or full-flexible? I read through the HADDOCK ab-ag tutorial and this part is not mentioned while still align quite well with the reference model. Just wondering if one should introduce such flexibility into CDRs of the Fab, what is the rule of thumb in terms of segment length, semi or full flexibility?

Many thanks for your time in advance!

2
  1. we want to model protein A-B complex, and both have a solved crystal structure respectively. Protein A structure is a tight conformation which is known to undergo drastic conformational change upon binding to its known partner. Now we have solid crosslinked restraint evidence connecting 3 out of 7 domains of Protein A with Protein B. Is there any suggestion regarding how to dock these 3 domain separately to Protein B and combine them together afterwards in 1:1 complex using HADDOCK? I have generated both DisVis-derived interface for those 3 domains-Protein B and corresponding distance restraints.

I would cut protein A into multiple domains and add connectivity restraints between those.
Also give full flexibility to the peptide regions connecting the domain (or even remove those if the amount of conformation change is too much).

The procedure is described in the following paper:

  1. I want to generate a antibody-antigen complex where antibody is modelled through AF2 multimer. DIsVis-derived interaction analysis provide a set of active residues. Do I need to specify which segment in CDRs to be semi-flexible or full-flexible? I read through the HADDOCK ab-ag tutorial and this part is not mentioned while still align quite well with the reference model. Just wondering if one should introduce such flexibility into CDRs of the Fab, what is the rule of thumb in terms of segment length, semi or full flexibility?

I would simply use the automatic flexibility option of HADDOCK. And you could consider using an ensemble of conformations for your antibody (and antigen) if you have those.

3

Thank you! I will take a closer look at the reference mentioned. For the ab-ag complex docking, I will go with the default setting first to see how it looks like. Unfortunately right now I don’t have any ensemble conformation data. Much appreciated for your valuable input!