I have a series of proteins that are all phylogenetically related. Only one of this proteins is currently X-ray crystallized.
Is it valid to modify the reference PDB file (I mean by hand) to match the mutated sequence?
And if so, is there any way to minimize the protein energy to find the most likely structure? (Trying to match as closely as possible the reference structure)
I would recommend you try the FoldX mutation engine. It is very easy to use and state of the art (it is the main competitor to expensive commercial software). It can do both of what you want - the mutation and energy optimisation. It may not work as well as molecular dynamics software, but for your simple task, it is a reasonable solution.
A simple command line can do it:
First, repair your structure (since, usually PDBs are not energetically optimised, so you could get better energy anyway without doing this):
FoldX --command=repairpdb --pdb=1AJY.pdb
Then, either use AlaScan or PositionScan to mutate. AlaScan only mutates to Alanine. PositionScan can mutate to any amino acid. More advanced mutations can be done with the BuildModel command.
FoldX --command=PositionScan --pdb=1AJY_Repair.pdb --output-file=1AJY_Repair_MA30d --out-pdb=false --positions=MA30d
Note: 'd' means try mutate the residue to 24 different amino acids; or, you can specify any of those 24, including the standard 20.
Which produces two files:
(It can also output the mutated pdb file by setting --out-pdb=true.)
Either of those files can be used to determine change in free energy upon mutation (ddg ~ delta delta g). The second file has additional information (the individual values used to calculate total change in free energy).
Note: The FoldX website has very good documentation for each FoldX command. You can find any additional information you require there.
Yes you can modify the reference PDB file and look for the changes and for this purpose you need visualizers. One among these is Chimera. You can easily carry out the energy minimization steps using Chimera, first mutate your PDB file manually and then provide it as an input for Chimera and the steps for doing energy minimization are fairly simple as given in Chimera docs here.
The previous answers suggest Chimera and FoldX, which require a license for commercial use.
Commercial users can use pdbfixer or a similar tool to perform mutations (using e.g. applyMutation in the API).
You can then use OpenMM to refine the mutated structure.
Granted, this requires some comfort writing Python scripts to use the APIs.
