By using the Rosetta application pmut_scan one can obtain a mutational
landscape of the ∆G for every possible mutation at each position.
In the case of enzymes, by repeating the same operation on the enzyme
without substrate docked, with substrate docked, with transition state
and with product one can find which residues are increase the reactivity.
This tool gets the output of different point-mutation scans, aligns them
into a single table and plots them. A few cases will fail entirely, hence
why the datasets are aligned by the tool.
As the energies are relative to wild type, it does not matter if in the
unbound state the ligand is not in the system or that in the transition
state has a strong constraint.
The quality of the results depends on the model, so it is important the
unconstrained scores form a typical Gibbs free energy curve, with negative
and positive ∆G‡
Due to the fact that preparing the models cannot be automated without
affecting the results, these are not done here.
If unfamiliar with Rosetta, feel free to contact Matteo Ferla (author),
who might be able to parameterise your ligands, prepare your models and
run them for you.
[Citation needed]Note that you have to cite the
Baker paper in addition to us...
The input files for this tool are the outputs from a pmut_scan run launched
along with relevant constraints and params files thusly:
pmut_scan_parallel -ex1 -ex1aro -ex2 -extrachi_cutoff 1 -DDG_cutoff 999 -mute basic core -s your_structure.pdb > scores.txt
is high because we want all mutants. Actually, despite the memory, adding
flag is also rather handy
for the analysis.
Provide one or more files (do not worry about header rows; the PDB numbering
will be used. No negative positions):