Suppose, I have the original PDB structure and I need to do certain mutations (may be, substituting amino acids) before performing MD Simulations or Normal Mode Analysis. Are there any reliable softwares? I know about SCWRL Algorithm (SCWRL4: http://dunbrack.fccc.edu/SCWRL3.php/), which has been extensively tested. Although SCWRL can select the optimal rotamer of the mutated residue, it doesn't model the change in backbone structure. If the nature of the amino-acid changes, there must be some local changes which will be propagated along the backbone and it will in turn affect the instantaneous positions of some residues/B-factor values etc. So, if I want to do some trajectory or network analysis with mutants, what method should I follow for creating the mutations??
You can use homology modelling programs, e.g. Swissmodel (https://swissmodel.expasy.org or accessed through DeepView https://spdbv.vital-it.ch, or modeller (https://salilab.org/modeller/) with Chimera as viewer (https://www.cgl.ucsf.edu/chimera/) which will do both the amino acid substitutions . Amino acid substitutions are easy, you can use delta-delta G prediction (rosetta, foldX) as indication whether backbone changes are needed to accomodated the substitutions. Rosetta Fixbb (https://www.rosettacommons.org/software) will also optimize the rotamers of the other residues in the protein to accomodate the mutation. Backbone adaptations can be introduced by energy minimization, in a first iteration putting constraints on all but the residues nearest to the location of the changes. Insertions and deletions are trickier, as they require loop modelling.
You can use homology modelling programs, e.g. Swissmodel (https://swissmodel.expasy.org or accessed through DeepView https://spdbv.vital-it.ch, or modeller (https://salilab.org/modeller/) with Chimera as viewer (https://www.cgl.ucsf.edu/chimera/) which will do both the amino acid substitutions . Amino acid substitutions are easy, you can use delta-delta G prediction (rosetta, foldX) as indication whether backbone changes are needed to accomodated the substitutions. Rosetta Fixbb (https://www.rosettacommons.org/software) will also optimize the rotamers of the other residues in the protein to accomodate the mutation. Backbone adaptations can be introduced by energy minimization, in a first iteration putting constraints on all but the residues nearest to the location of the changes. Insertions and deletions are trickier, as they require loop modelling.
Hello Olivia Debnath,
I would recommend you to use PyMOL which is a very effective Molecular Visualization software for performing In-Silico Mutagenesis.
Please follow the below link for step by step procedure of performing Mutagenesis and for the explanation of their respective color codes.
https://pymolwiki.org/index.php/Mutagenesis
Best,
Rahul Ravichandran.
Hello Rahul,
I have already used PyMol for the mutagenesis purpose but while checking the PDB file, I came across that it is not changing the x,y,z coordinates for the substituted amino acid. Moreover, the B-factor value is becoming zero for the mutated residue which shouldn't be the case. I did it for the 'Backbone Dependence Rotamers'. I am scared it will give me a wrong trajectory after MD.
No, no program assigns B-factors to modelled residues - B-factors are a measure of uncertainty of the coordinates, dependent on the experimental procedure used to determine the structure. Local B-factors may be influenced by crystal contacts.
In PyMol, you can manually click through the main rotamers of an amino acid to select one that produces the fewest clashes, however, it does not optimize the rotamer choice. Homology modelling programs handle rotamer choice differently - some use the most common rotamer for the backbone conformations, others the one closest to the rotamer of the original amino acid in this position. Rosetta does a repacking of all side chains.
PyMol does not do any energy minimization, therefore the coordinates of the backbone and Cbeta atoms do not change - as long as you have no major clashes, the energy minimization you do when preparing the structure for MD will take care of that. If you have bad clashes, you need to resolve them before doing an unconstrained energy minimization, otherwise, the minimization can badly distort your structure.
Beware: some modelling programs will set the occupancy of modelled residues - if this causes a problem, either correct them to 1.00, or set the subsequently used to ignor occupancy settings
Hello Olivia
As suggested by Annemarie, modeling the protein from the scratch, i.e, sequence, after inserting the mutation in particular residue position, is the best possible method to overcome all the steric clashes and accommodate backbone changes in the protein.
In our lab, we also frequency perform mutational studies computationally,
but we generally mutate the residues using Chimera or Schrodinger prime module to mutate the residues.
Schrodinger Prime module, indicate the clashes or local changes in protein after mutations (side chain & backbone), which can be corrected easily.
However, earlier we used to minimize the mutated protein using the Conjugate gradient energy minimization method for around 50000 steps in chimera tool. Mostly, it gave quite accurate predictions as most of the computational predictions complemented the Wet lab data that we obtain.
So, that might be helpful to you.
Alternatively, if you have good computational resources and time you can perform MD studies for 5-10ns using any MD package like Gromacs, Amber, NAMD or Desmond. More parameters you use with proper knowledge like Salt concentration, ions presence, cofactors, etc, more the accuracy.Although default parameters for the protein can be used.
After MD, the protein mostly acquire a more stabilized conformation.
Hope, it will work for you.
regards
Tanuj
You can do an MD after introducing mutation followed by an FEP calculations.
- Badges
- Science topic