I selected a protein from the protein data bank for my MD simulation study. But the selected protein is not the wild type since it has mutations and missing residues. I was unable to find the wild type as well. So how can I get or modify the protein structure without mutations and including missing residues? And if I performed MD simulation directly using the mutated protein will that give wrong result because of the mutations? If I have to remove mutations and add missing residues what software and server are freely available?
HI @PriZarah
And if I performed MD simulation directly using the mutated protein will that give wrong result because of the mutations?
What are you trying to do? Are you trying to replicate experimental data with the mutation? Then maybe keeping it is the best option. Do you want to compare behavior with/without mutation? Then you will need both. Do you care about the WT but the mutation is far away from relevant parts of the protein system? Then maybe you can keep, although changing it would be the best option, in my opinion at least. This is a very general question that can’t be answered here, but you should consider yourself and discuss with your supervisor/someone that is interested in the results and outline properly the research question BEFORE running the simulation.
I was unable to find the wild type as well.
I guess here you refer to the crystal WT? But you know the full sequence, so you should be able to backmutate if needed.
If I have to remove mutations and add missing residues what software and server are freely available?
There are several free tools that can help you in this. Pymol and ChimeraX are both free and can help you change rotamers, point-mutate, etc. CHARMM-GUI is a web-server that is also able to manipulate PDBs. If there are missing sections, e.g. un-modelled loops, then you can use tools like MODELLER or AlphaFold. For optimization, loop filling, mutations and similar also the Protein preparation suite of Maestro is a good tool, but very very expensive, and you can do the same with the other tools I linked. It’s just worth citing Maestro as maybe your institution might have some sort of (partial) free license for you to use, and is very relevant if you need to dock ligands as well.
1 Like
In addition to the excellent reply by @obZehn, I’ll add a couple of points:
Think of designing the simulation (or set of simulations) as an experiment with a hypothesis to test, what is your research question in this case?
Fundamentally, mutations could affect the stability and function of the protein. Check why was the structure mutated in the paper where the structure was published, is it a stabilizing mutation so the protein could be crystallized, for example? In addition, even mutations far from ligand binding or functional sites could affect function via allosteric effects. So unless your research question requires the mutation to be present, I would highly recommend to back-mutate the structure back to WT before running the simulation