WHAM is used extensively to calculate the free energy of a system along a properly chosen reaction coordinate in molecular dynamics simulations or monte carlo simulations. To be able to use WHAM you need to perform multiple simulations that cover the entire reaction coordinate to sample possible configurations. The choice of reaction coordinate constitutes a length discussion and is a topic of ongoing debate. So I encourage the reader to look it up elsewhere. After you have chosen a reaction coordinate and performed Umbrella Sampling or Free Energy Perturbation calculations, histograms corresponding to every window need to be extracted.
For the simple case of dragging a methane molecule from water into a lipid bilayer and out into water again, z axis is a good reaction coordinate if the lipid bilayer is on the x-y plane. With this set up, the z axis is split into discrete windows and the center of mass (COM) of the methane molecule is constrained (usually harmonic) to the center of every window. The length of windows and force constant of the constraint are chose such that the distribution of the z coordinate of the COM of methane overlaps with the distribution of the adjacent windows. This is a necessary prerequisite for using WHAM. Following the figure below, for a 60A long reaction coordinate a choice of 121 windows (each 0.5A and including both ends) will be plenty to ensure overlap of z coordinate distributions.
Collect z coordinate distributions and put them 121 files labeled window-001 to window-121.
Create a new file input.in that contains the following lines;
./window-001 -30.0 xx
./window-002 -29.5 xx
...
./window-120 +29.5 xx
./window-121 +30.0 xx
Here the first column is the file name, second is the mean value of that window and the third column needs to be filled with an integer but is irrelevant and not used by Alan Grossfield's implmentation of WHAM. Then run the following command
./wham 1 11 10 0.001 300 0 input.in freefile
where the inputs are
./wham hist.min hist.max num.bins tol Temp numpad infile outfile
The output contains a plot for the potential of mean force (PMF). Pay particular attention to units since they depend on your input. WHAM's documentation can be found here [http://membrane.urmc.rochester.edu/sites/default/files/wham/doc.html].
For the simple case of dragging a methane molecule from water into a lipid bilayer and out into water again, z axis is a good reaction coordinate if the lipid bilayer is on the x-y plane. With this set up, the z axis is split into discrete windows and the center of mass (COM) of the methane molecule is constrained (usually harmonic) to the center of every window. The length of windows and force constant of the constraint are chose such that the distribution of the z coordinate of the COM of methane overlaps with the distribution of the adjacent windows. This is a necessary prerequisite for using WHAM. Following the figure below, for a 60A long reaction coordinate a choice of 121 windows (each 0.5A and including both ends) will be plenty to ensure overlap of z coordinate distributions.
Collect z coordinate distributions and put them 121 files labeled window-001 to window-121.
Create a new file input.in that contains the following lines;
./window-001 -30.0 xx
./window-002 -29.5 xx
...
./window-120 +29.5 xx
./window-121 +30.0 xx
Here the first column is the file name, second is the mean value of that window and the third column needs to be filled with an integer but is irrelevant and not used by Alan Grossfield's implmentation of WHAM. Then run the following command
./wham 1 11 10 0.001 300 0 input.in freefile
where the inputs are
./wham hist.min hist.max num.bins tol Temp numpad
 |
| Figure Source: http://www.utdallas.edu/~son051000/comp/FreeE.pdf |
The output contains a plot for the potential of mean force (PMF). Pay particular attention to units since they depend on your input. WHAM's documentation can be found here [http://membrane.urmc.rochester.edu/sites/default/files/wham/doc.html].
No comments:
Post a Comment