CHARMM Development Project Forums User Discussion & Questions Parameter Set Discussion Ligand-Protein Interaction Problems

Another problem happens now...

I was trying to read the ligand MTX with DHFR protein together to study interactions.

I am giving here *.inp file, the *.out file, *.rtf and *.par file of the ligand. Download the zip attachment and unzip it to have the files.

It seems that CHARMM is creating an unwanted and unnecessary bond between the ligand and the last residue of the protein.

Tell me how to overcome this problem...

I thought you wanted a bond between the protein and the ligand?
You have to tell CHARMM which atoms should be bonded, either by including this information in your MTX residue (probably the easiest way), or by using a PRES, which can also be used to delete unwanted bonds if there is no way to keep them from being formed. The only bond that CHARMM would form "automatically" between the protein and your ligand is between the C-terminal C of the protein and an atom named N in your ligand.
You have started a new thread, so I cannot easily check your previous posts.

Your atom types and dihedral parameters look like they are derived from the MMFF force field. Also, the charges are quite different from CHARMM charges; they may be MMFF charges. This is not expected to work well with a protein represented by the CHARMM force field. Even more troubling, the L-J parameters appear to be derived from CHARMM or OPLS. Class I additive force fields strongly rely on compensation of errors between the charges and L-J parameters, and between these nonbonded interactions and the dihedral parameters. Therefore, using L-J parameters from one force field with charges and dihedrals from a different force field is asking for trouble with the nonbonded interactions and conformational preference (respectively). And MMFF is very much different from CHARMM in the way it represents VdW interactions.

@Kenno,

But what can I do now? I have used the SwissParam server to generate the rtf and par files. I have the molecule in mol2 format and I cannot afford to write my own rtf and par files now.

If I have to write my own rtf and par files now, my project will be delayed.

Please suggest me a way out so that I can have the rtf and parameter files in my hand (in proper format).

The ParamChem site (the only such site recommended for adding new molecules designed to be fully compatible with the CHARMM force field) accepts molecules in mol2 format.

Okay...

I have generated the stream file using the ParamChem server of another ligand. The energy minimization and even the molecular dynamics simulation of only the ligand are performed successfully.

Now when I want to associate it with its receptor (sequence read from downloaded pdb), CHARMM is showing again that some angle parameters are missing in the protein. The protein has no special residue or modified/synthetic amino acid residue.

Any clue on how we can overcome this problem?

In the case of my MTX ligand, ParamChem is showing "Amide Bond not Supported" error...

How this problem should be addressed?

There is not enough information to help with your missing angles. At the very least, you have to show us which ones are missing. Also, please clarify whether you want your ligand to be covalently bound to your protein or not.

As for your MTX ligand, there is no "Amide Bond not Supported" error anywhere in my source. You must have misread the message. Towards solving the problem, reading the FAQ would be a good start.

Edit: also, generally spoken, please take note of the Parameter Set Discussion forum guidelines.

attype warning: amide base not supported;
skipped molecule.

This is the error...

Error Information
Processing uploaded File resulted in a few errors and/or warnings:
attype warning: amide base not supported;
skipped molecule.
Output Information
Error processing uploaded input file. Please check the error information.

The MTX ligand has been extracted from the crystallographic model...