[gmx-users] questions about umbrella sampling

[chris.neale at utoronto.ca]

Dear R-X Gu,

I think that you didn't get a reply because you are clearly using  
gromacs version 3.x and many of us now work with gromacs 4.x. (Note  
that the .ppa file is gone in gromacs version 4.x and the pull  
parameters now go into the regular .mdp file). If there is some reason  
that you need to use version 3.x then I am familiar with that and I  
can help you with it. But it's a waste of time if you're going to move  
to version 4.x soon.

Still, I can answer your question generally right now. First, you are  
correct that you don't need to set the system up with zero restraint  
energy (that is what you would have if the displacement from desired  
distance was exactly 0.0). get it close though... if it is too far  
then the initial forces will be large and can introduce unintended and  
not always easy to notice biases or structural deformations.

Second, to calculate the center of mass displacement, you can use g_dist.

Your questions indicate that you are new to gromacs. So let me provide  
you with some unrequested advice. First, double posting to a mailing  
list is never useful. We all saw your first post. If you don't get a  
reply then learn some on your own and when you post again you can  
include this new information to say "I tried X and Y, but they lead to  
Z, which is not what I intended...". Second, free energy calculations  
are difficult and not recommended for beginners. My intention here is  
not to dissuade you from doing them, but to suggest that you spend a  
few months doing some regular simulations first. Third, if you want to  
do umbrella sampling free energy calculations, find a tutorial and  
start there. This is a good one:  
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/umbrella/index.html but I am sure that there are lots more. In fact, there's a gromacs page for tutorials on many topics:  
http://www.gromacs.org/Documentation/Tutorials

Chris.

-- original message --

Dear All:

I am using umbrella sampling method to calculate potential of mean
force and I encountered the following questions:
1, at the *.ppa file, I need to incorporate a parameter indicating the
"the position of the pull group will restraint to, relative to the
reference group" (i.e., the "pos1" parameter).

To my understanding, in the initial structure, the pull group is not
need to at the exact position defined by the aformationed parameter
("pos1" parameter) in the .ppa file. I just need to put the pull group
near that position, am I right?

2, Since I need to define the position of the pull group relative to
the reference group. I need to calculate the mass center of the
reference group as well as that of the pull group. Does there any
tools in GROMACS to calculate their mass center?

Thank you very much for your time and your kindness!

Best Regards

R-X Gu