[gmx-users] Umbrella sampling and free energy estimation

[James Starlight]

Dear Gromacs Users!


Recently I've completed Umbrella sampling tutorial presented on the
Justin's web. I'd like to use that teqnique as the engine for
enhansing of conformational sampling of some monomeric proteins
(starting with the simplest example of apo-calmodulin) as well as
estimation of the free energy of apo form of my protein ( something
like free energy surface)/ So I have some methodolofical questrions
about emplementration of the umbrella sampling to the  such tasks.
1- Does it possible to obtain representation of the distribution of
conformational free energy of my protein by means of that method ? (
estimation of what conformations are most hightly-populated from
overal protein ensemble )

2- Also I wounder to know about pulling algorithm. Might I use some
eigenvector (or the combination of selected eigenvectors) extracted by
the covariance analysis of the equiliubrium trajectory or by the PCA
from the experimental data-set of the simulated protein ?
What changes to the pulling algorithm should I add for such simulation?
Also I found in that settings 2 groups- where  pull_group1  is what
I'd like to pull.
pull_group0     = Chain_B
pull_group1     = Chain_A

 If I have just one group ( protein itself) which I'd like to pujll
along selected coordinate what I should specify as reference
(pull_group0) ?


Thanks for help,


James