Hi everyone!
I checked the CHARMM36 port for GROMACS, and after opening the ions.itp file, I did not find fluoride parameters inside.
Are there no NBFIX parameters for fluoride, or are they not portable for GROMACS?
I’d like to simulate systems with high salt concentration (up to 4M of KCl and KF) using CHARMM NBFIX parameters.
Thank you to everyone who could help me!
CHARMM does not support F-.
Thank you, Justin.
Dear Dr Lemkul,
I am new to gromacs and all of your tutorials have helped me perform MD simulations easily. But now I need to modify quite a few residues, changing ser/thr/tyr into their phosphrilated forms. I know that the charmm ports contain all the information needed, but I don’t how to apply them. Is there a tutorial somewhere that can walk me through this protocol step-by-step?
Thanks a lot
Mario
The residue names are the same as the corresponding patches in CHARMM, so you need to change them in the PDB file. SP1 = pSer(-1), SP2 = pSer(-2), THP1/THP2 for pThr, TP1/TP2 for pTyr.
Thanks a lot for the swift reply. I again phosphorylated a small peptide at Tyr by using the pymol PTM and charmm-gui and, though changing PTR or TYR into TP2
the response is the following (charmm36-jul2022.ff was added in the working directory)
Fatal error:
The residues in the chain ALA320–ILE325 do not have a consistent type. The
first residue has type ‘Protein’, while residue TP2323 is of type ‘Other’.
Either there is a mistake in your chain, or it includes nonstandard residue
names that have not yet been added to the residuetypes.dat file in the GROMACS
library directory.
I checked the residuetypes.dat file in the charmm36-jul2022.ff, where there are TP1 and TP2, but not SP1/SP2 or THP1/THP2. Should I add these latter by myself? On the other hand, TP1/TP2, TPH1/TPH2 and TP1/TP2 are shown in their full form in the aminoacids.rtp file.
Only when I mutate TYR into TP2 by using charmm-gui until the second step of the forcefield converter, and use the new pdb file for the pdb2gmx step, a topology file comes in. This does not occur if I use the pdb file generated by the first step in the charmm-gui.
Just notice a line of the first pdb file is
ATOM 40 N TP2 P 323 11.095 -79.662 -25.587 1.00 20.76 PROA
and after the FF converter is
ATOM 40 N TP2 323 11.095 -79.662 -25.587 1.00 0.00 PROA N
Not a big difference, but why does the former not work and dos the fatal error turns up?
Sorry to annoy you, but I really love gromacs, and charmm is easy to use for me for complexes, but now I would like to dive into post-translational modifications, especially phophorylation (it is my field of interest in biochemistry) and sometimes things are a bit hard for me to understand even if I spend a lot of time on reading and trying to learn.
Thanks a lot again
Warm regards
Mario
It is not possible to anticipate all residue names that might be considered part of one biomolecule or another, so if you’re using TP2 as an amino acid name, you have to add it to residuetypes.dat as a Protein residue. You can do this in a copy of residuetypes.dat in the working directory, which will be read before the one in $GMXLIB.
Thanks again. I figured out how to modify the files in the .ff folder. Everything seems to work fine. I had to put together Protein, SP2, TP2 and LIG by writing Protein_SP2_TP2_LIG (tc-groups) in the nvt.mdp, npt.mdp and md.mdp, and not only Protein_LIG. Is it correct or may I have issues?
It will be a lot more convenient in general if you define the modified residues as Protein in residuetypes.dat, that way you don’t have to keep making special index groups for tc-grps or every other analysis tool to understand that they’re protein.
Thats what I did by adding SP1/2 etc to the residuetypes.dat file defining them as Protein. But these residues still appear in the index file. Any further suggestions?
Really grateful to you
Mario
PS. By the way, in a few minutes I’ll analyze the MD simulation with two phosphorylated residues and a ligand to check whether there are differences between phosphorylated and non-phosphorylated complexes.
There must be another database file, maybe selections.dat if memory serves, that might have to be updated for default selections. In any case, no harm in what you’re doing, it just makes it more tedious for generating groups.
Although it is a bit cumbersome, adding ProteinSP2_TP2_LIG worked for me. There is no selection.dat or whatever in the charm36-jul2022.ff folder.
Thanks a lot for your support, I made it
Mario
Dear Dr Lemkul,
I wish I could run an MD simulation of a protein kinase with ATP-2Mg2+ within its active site in the presence of an ATP-competitive ligand to assess whether this latter can remove ATP from its binding pocket. I managed to generate the topology of a PKA- ATP-2Mg2+ complex from the pdb repository (PDB: 4DH3) by using charmm36-jul2022.
My question is whether I can continue by following the usual protocol as described at mdtutorials.com, or I should extract ATP (along with the 2 Mg2+ ions?) out of the pdb file and generate a new topology as in the case of any ligands , especially in light of the position restraints I have to add to the topology itself. It goes without saying that pdb2gmx generated the posre_*.itp files for all the species of the original pdb. Do you suggest me leaving ATP in the current topology or manipulate it as a " ligand?
Thanks in advance for any suggestions
Mario
The force field already has everything you need for ATP and Mg2+. Generate the whole system topology with pdb2gmx.
Sorry to annoy you again. I have a benzotriazole with two bromines and two iodines bound to the benzene ring. The bonds appear in the .pdb and .mol2 format, but after parametrization with cgenff or charmm-gui, both halogens are misrepresented as “dots” (upper picture attached, lig_ini.pdb) or weird bonds occur (lower picture attached, ligandrm.pdb) respectively.
attached)
. I know that you published a paper on halogen parametrization, but maybe I am a bit slow (I am not in the position using gaussian and other softwares). Is there a method to overcome this issue? Thanks a lot for any suggestions
Mario
It’s a visualization issue. The bonds are long and the software doesn’t know what to do with them. It would be better to post specific help requests like this in the user forum rather than the third-party tools announcement forum.