GROMACS version:2020.1-Ubuntu-2020.1-1
GROMACS modification: No
I am trying to simulate a Protein-Ligand dimer (PDB: 1g6n), the CAP-cAMP complex. I am following the standard protein-ligand complex tutorial for the same. Following the tutorial, I separated my ligand (A monomer instance of cAMP, not both the cAMPs) and protein into separate .pdb files. Next, I added H-bonds to the ligand in Avogadro2 and imported it to a mol2 format using OpenBabel. Parametrization using CGenFF resulted in poor penalty values, and I tried to optimize parameters using FFParam, which is still ongoing. However, I decided to try running a simulation using parameter optimization provided by SwissParam. Using SwissParam, I obtained the cmp.itp file, which I referenced into the topol.top file by the #include command. I also added the entry LIG (my molecule is universally referenced as LIG) into the [ molecules ] directory and set the quantity to 2. (Following advice on my previous forum post here).
Now, the tutorial instructs one to make the complex.gro and topol.top files consistent with each other by adding the ligand atoms from `ligand. gro ’ and updating the number of atoms.
However, I am confused what exactly my complex.gro file should look like since I want two instances of the ligand in my simulation. I tried copy-pasting the ligand twice in the complex.gro file, i.e. same coordinates for two cAMP molecules but different instances. This did not work out and resulted in a segmentation fault during EM, which I think was because two instances of the ligand should never overlap. Hence, I would appreciate any advice on how I can proceed with the simulation setup here. I have attached Pastebin links of all relevant files below (can be viewed on browsers without downloading the file).
topol.top
cmp.gro (converted from .mol2 using the gmx editconf command).
cmp.itp
complex.gro ( truncated, showing relevant parts)
No, your two ligand molecules definitely shouldn’t overlap. Just move one of them away, either manually in VMD (there’s the move \rightarrow representation or move \rightarrow molecule mode there), or anything like gmx editconf, or even manually adding a number to a selected coordinate.
1 Like
Hi Milosz,
I had a follow-up query. Would placing the second ligand molecule anywhere within the complex work, or should it be placed as close as possible to it’s ideal position (in the original pdb file)?
Now that I looked at the structure you provided, I noticed I don’t understand why you want one protein chain but two ligand chains (at least that’s how I got it). What’s the purpose of the other ligand molecule? Or am I missing something?
Hi Milosz,
The complex.gro file I provided is truncated (not the whole picture). I am indeed using both protein chains, and the reason I need two ligand molecules is that in the CAP-cAMP system, there are 2 cAMP (ligand) molecules.
I have added the complete complex.gro file here. In this complex, I moved the cAMP(ligand) molecules using Pymol to make them lie inside the protein dimer. Since in Jake’s tutorial, position restraints are added on the protein+ligand or ligand, I was concerned if the position of the cAMP (ligand) in the complex.gro file matters, or would it find its optimal position during the course of the production run? (I thought it wouldn’t due to position restraints).
But then you have the original positions of both ligands in the original PDB entry. Is the problem just that you need to protonate them to match your separately prepared ligand topology? If so, the easiest thing to do is to align both individual ligands to the original PDB’s heavy atoms, and there are quite a few ways to do that (I like the flexibility of MDTraj, but VMD would do too).
Although cAMP doesn’t have a separate magnesium ion so shouldn’t be as sensitive as ATP to small translations, starting from a bound geometry that’s as close to experimental as possible should be better for the stability of your simulated complex.
Hi Milosz,
Yes, the problem was that I needed to protonate them separately to match the prepared ligand topology. I will look into aligning the separated ligands to the original co-ordinates, thank you!