Hi everyone,
I see some strange behavior of enzyme design with ligand.
1) I have created a library of rotamers, set path to this library in params file (by PDB_ROTAMERS rotlib.pdb) and run enzyme design. But there is still only translation and rotation of rigid body in the active site and no using of the rotamer library. Only when I start with ligand structure with some clashes in the protein pdb file I can see optimization of of torsion angles of the ligand. Can you please explain how it works?
2) I have a problem with cst file. Even when I set the force constant to really high value (CONSTRAINT:: distanceAB: 2.4 0.3 999.0 0 1) I'm not able to keep the defined atoms (C and O) closer then 3 A.
I would be very grateful if someone could help,
Dave
If you have the PDB_ROTAMERS line set correctly, the rotamers therein should be used for packing. There should be information about the loading of the rotamers in the "core.pack.dunbrack.SingleLigandRotamerLibrary" tracer on standard output. Are you getting any messages from that tracer?
With your constraint, you're not setting the constraint as "covalent". Put a "1" in the periodicity column of the distance line to get a covalent constraint, which will turn off atom/atom repulsion for the two constrained atoms, an allow them to get in closer contact. (https://www.rosettacommons.org/docs/latest/match-cstfile-format.html)
You're rotamer issue might be related to your constraint issue. If you superimpose a rotamer, and that rotamer greatly moves the location of the constrained atom, the energy of that rotamer will be bad, so the packer might not accept that rotamer substitution. What you may want to do is look at the "neighbor"/NBR atom for your ligand. When the rotamers are overlaid, they'll be overlaid and oriented on the neighbor atom. Depending on how you want your ligand flexibility to be modeled, you may need to change your neighbor atom. (You can manually edit the params file, but if you do, you'll want to change the NBR_RADIUS as well - all heavy atoms for any rotamer should fall within this distance to the neighbor atom.)