Trimer orientation moves during ligand dock


I am docking a polysaccharide onto a trimeric protein structure. I have merged the trimer into a single chain (e.g. promoter 1 - res 1-130; proteome 2 - res 201-330, etc) and the ligand is docked across two of the promoters. However, after the HADDOCK refinement of the docked complex the relative orientation the trimer chains has changed. Based on subsequent MD simulations this does not appear correct and the original trimer orientation is re-established, which results in half of the polysaccharide being displaced. Is there a way to either still use the trimer within one chain ID or to use the trimer with different chain IDs, but fix their relative orientations.

Many thanks for any help


Are you running on the webserver or locally?

If on the server, which interface did you use to submit your data?
The server automatically detects chain breaks and defines restraints between the chains. So this should in principle not happen or something is wrong.

Thanks for the quick response. I am docking heparin and need to run locally so I can define restraints to individual hep sugars. Should I use several internal distance restraints within the protein to lock it in place then?

Many thanks


Simply measure a few distances between e.g. CA atoms of your three chains and input those are unambiguous restraints.

Our haddock-tools repo does contain a script for this purpose: