You are here

ddg with ligands

5 posts / 0 new
Last post
ddg with ligands

Could someone please recommend a way to include ligands in a ddg calculation (using ddg_monomer or otherwise)? I'd like to estimate the effect of mutations to the protein domain on the stability of the interface between said protein domain and the enzyme substrates.

Thank you in advance for any response!

Post Situation: 
Mon, 2011-12-05 14:32

If you're lucky, you should just be able to just use the existing application with an input file with a ligand in it and provide the appropriate params file with the -extra_res_fa flag. That's not always the case, though, as often protocols written for monomeric proteins (or even protein-protein interactions) will choke when given multiple chains or non-protein residues.

An alternative if that doesn't work is to put something together with RosettaScripts. We've had some moderate success using the FilterScan filter to evaluate mutations. The protocols haven't been worked out, though, so you'll likely have to play around with how best to do the relaxing/repacking/minimizations/etc.

Mon, 2011-12-05 15:16

If you just want to calculate the ddg of a ligand you do it using the InterfaceScoreCalculator mover, with a script line like this:

<InterfaceScoreCalculator name="add_scores" chains="X" scorefxn="score12" native="native_structure.pdb"/> </code>

Where "X" can be replaced with the chain ID of your ligand (it has to be a different chain than the protein), "score12" can be replaced with your score function of choice, and "native_structure.pdb" is the path to a native pdb for RMS calculations (this is optional). total ddg is calculated as "interface_delta_<chain>" and will appear at the end of your pdb file, or in the SCORES: line in a silent file.

This mover also outputs the ddg split into component score terms, these are tagged in the form if_<chain>_<score_term>

Tue, 2011-12-06 12:28

Thank you for your responses. ddg_monomer appears to have accepted the -extra_res_fa flag. The only adjustment I had to make was to rename an atom in the ligand CA because of the way the constraints are defined (following the example in the manual). Any idea if that is reasonable? Or if it might matter which atom is involved in the constraints?

Thu, 2011-12-08 14:27

If all the CA are being used for are the AtomPair constraints to keep the backbone from moving too much, you should be fine.

Which atom is renamed "CA" would make a difference if you have a floppy/flexible ligand. The atom which is renamed CA would be restrained in its position, so you probably would want to put it near the "core" of the ligand, in a place which you wouldn't expect to move much from mutant to mutant.

I'm not familiar with the ddg_monomer application, but if I'm reading the documentation correctly, you could possibly even remove those lines referencing the ligand CA atom from the constraint file, if you don't want to constrain the ligand position.

Thu, 2011-12-08 18:01