GROMACS version: 2018.6
GROMACS modification: No
Here post your question
Hi, I would like to analyze a compound bound with a nuclear receptor protein by MD simulation using GROMACS software; I had successfully do it with a ligand-protein complex. If I want to add a cofactor pepetide which is the protein’s cofactor in order to further analyze the binding effects. How can I do it with 1 protein, 1 ligand, and 1 cofactor peptide. Until now, I only know to set two groups in parameters for simulations. One is protein and the other is ligand. I do not know how to add a peptide to do MD simulation with the protein-ligand complex. Can anyone help me?
Thank you
The approach is no different from any protein-ligand complex. The peptide “ligand” is just another protein chain and pdb2gmx can natively write topologies for multiple chains.
Thank you for your answer @jalemkul
It is very available for my studying.
According to your answer,
it is available to add the peptide to its bound protein to further generate a new protein file (protein group) for generating a new protein-ligand complex for performing MD simulation. is it right?
I have a new question.
If the bound ligand (compound) caused the protein’s conformation change a lot, is it possible to display a situation that the peptide leaves its bound protein after MD simulation, moved away from the bound protein or bound more closer ?
I worried that the properties of the new generated protein file “peptide + protein” (protein group) may be changed. Whatever, the peptide was not belonged to the part of the protein. I worry to combine both of them to generate a new protein file to perform MD simulation. Is it necessary to be considered for that or not ?
Anything consisting of amino acids will be recognized as protein by all GROMACS utilities. You don’t have to do anything for this to work.
I’m not sure I fully understand the question. But if you want to model relocation of a bound peptide or its dissociation, I doubt the time scales accessible to conventional MD simulations will allow you to do this. You may need an enhanced sampling method.
I would like to analyze the ligand (ligand group) binding behavior on the protein.
I also want to know whether the ligand causes the protein’s conformation changes to make the location change of the peptide or not.
Can I do these works according to the traditional protein - ligand MD simulation according to your tutorial ?
Do I need to modify the protocol of the tutorial according to “enhanced sampling method” you mentioned ?
If it is necessary to do with modified protocol, how can I do it ?
If you are truly interested in studying the binding process, this is a massive project and there is a large body of literature on applicable techniques. Probably some kind of metadynamics or machine-learning biasing potential is what you need, but that is far more advanced than anything in my tutorials.
What kinds of situations that I can add the peptide to its bound protein to generate a new protein file (protein group) for performing MD simulation ?
At the first time, your answer is " The approach is no different from any protein-ligand complex. The peptide “ligand” is just another protein chain and pdb2gmx can natively write topologies for multiple chains." Why now answer is no ? is it different? or is it limited in some situations?
Your question changed over time. Performing an MD simulation of a protein:ligand complex and a protein:peptide:ligand is not functionally different. But if you want to do something complicated like simulating unbiased binding, you need to do something other than brute force MD. Generating the topology, however, is no different. The protein and peptide topology come from pdb2gmx. The ligand comes from an external tool and its topology is integrated into that of the system. What you do from there varies depending upon the application.
In my approach, I would like to know the binding stability of protein:peptide:ligand complex.
Whether I can do RMSD or RMSF analysis to analyze the binding stability of protein:peptide:ligand complex after performing a MD simulation with the complex according to your tutorial ?
Are these trajectory analyses (RMSD, RMSF, Hbond, Rg) suitable to be performed ?
The analysis needs to be motivated by whatever the relevant dynamics are in the system. What you have listed are boilerplate analyses that may or may not be useful. Try them and see. But your most valuable tool is your eyes. What you see in the trajectory can and should guide how you quantify the behaviors you observe.
You are right! Some case of MD simulation. I found that the fluctuation of helix displayed a significant moving that was observed by my eyes. However, it was not analyzed by RMSF analysis. Do you know why ?
Let’s me try it. I hope those trajectory analysis methods (RMSD, RMSF, Rg, Hbond) can help me analyze the binding stability of protein:peptide:ligand complex after performing a MD simulation according to your tutorial.
I tried to do pdb2gmx, but produced 2 types of posre*.itp file and 2 types of topol*.itp file. I do not know how to do next steps. Besides, a error message (invalid order for directive moleculetype) was happened, shown as below:
protein had two chain, A and I chain, shown as below,
ATOM 1795 CA GLY A 368 59.033 58.052 15.305 1.00 97.56 C
ATOM 1796 C GLY A 368 58.593 58.349 13.883 1.00110.38 C
ATOM 1797 O GLY A 368 58.377 59.536 13.560 1.00111.51 O
ATOM 1798 OXT GLY A 368 58.471 57.398 13.081 1.00109.05 O1-
TER
ATOM 1799 N LYS I 688 59.481 63.655 19.378 1.00139.48 N1+
ATOM 1800 CA LYS I 688 60.143 64.297 20.548 1.00146.68 C
ATOM 1801 C LYS I 688 59.263 65.370 21.186 1.00148.41 C
ATOM 1802 O LYS I 688 59.573 66.561 21.127 1.00150.33 O
when I do pdb2gmx, it produced the following files,
protein_processed.gro
posre_Protein_chain_A.itp
posre_Protein_chain_I.itp
topol.top
topol_Protein_chain_A.itp
topol_Protein_chain_I.itp
,and I make a copy of protein_processed.gro to be a complex.gro;
At a time before, I moved the all “;include” to the top before the [moleculetype]
However, in this case, I don’t know how to perform and how to solve the error. How to moved the ";include " to a suitable location?
