Re: Colvars and distance restraint for umbrella sampling

From: Gumbart, James C (gumbart_at_physics.gatech.edu)
Date: Wed Oct 17 2018 - 10:49:11 CDT

I think the forceconstant you are using is way too high. I believe the scaling by the width goes in the other direction, i.e., you are using a force constant of 50,000 kcal/mol*A^2. Can you please check your logfile to confirm? It will say something like “Force constant for ___ rescaled to…”. I could be wrong.

As for checking the overlap, you can plot the traj files directly and then make histograms from them (xmgrace, for example, can easily do this). The tutorial also gives ways to determine this, and other things to check, on pages 23-25.

Best,
JC

On Oct 17, 2018, at 11:41 AM, Udaya Dahal <dahal.udaya_at_gmail.com<mailto:dahal.udaya_at_gmail.com>> wrote:

Hi JC,

I didn’t run the tutorial itself but i am able to run the US now . The only difference I noticed is the force constant used in the tutorial is not able to constraint my peptide in the target center. I significantly increased the value which is now restraining the peptide around center of mass. Currently I have 20 windows of separation width of 1Angstrom and I am using 0.1 in the colvar width. How do we know if the neighboring windows overlap and how do we determine what force constant to use? Here is my recent colvar script

colvarsTrajFrequency 20
colvarsRestartFrequency 5000


colvar {
   name Translocation

   width 0.1

   lowerboundary 40.0
   upperboundary 100.0

   lowerwallconstant 100.0
   upperwallconstant 100.0

   distance {
      group1 { atomnumbers {7301 7302 7303 7304 7305 7306 7307 7308 7309 7310 7311 7312 7313 7314 7315 7316 7317 7318 7319 7320 7321 7322 7323 7324 7325 7326 7327 7328 7329 7330 7331 7332 7333 7334 7335 7336 7337 7338 7339 7340 7341 7342 7343 7344 7345 7346 7347 7348 7349 7350 7351 7352 7353 7354 7355 7356 7357 7358 7359 7360 7361 7362 7363 7364 7365 7366 7367 7368 7369 7370 7371 7372 7373 7374 7375 7376 7377 7378 7379 7380 7381 7382 7383 7384 7385 7386 7387 7388 7389 7390 7391 7392 7393 7394 7395 7396 7397 7398 7401 } }
      group2 { atomnumbers { 1910 }}
      }
   }

harmonic {
colvars Translocation
        centers 11.5
        forceConstant 500;# 500 * 0.1^2 = 5kcal/mol
Regards,
On Oct 17, 2018, at 11:27 AM, Gumbart, James C <gumbart_at_physics.gatech.edu<mailto:gumbart_at_physics.gatech.edu>> wrote:

If you find something taken directly from that tutorial no longer works, please let me know.

You don’t say what error you got, so it’s difficult to make suggestions (other than what Giacomo noticed). Also, regarding center of mass, I think all multi-atom colvar components use the center of mass (or geometry?) by design.

Best,
JC

On Oct 16, 2018, at 8:41 PM, Udaya Dahal <dahal.udaya_at_gmail.com<mailto:dahal.udaya_at_gmail.com>> wrote:

Thank you Giacomo for the quick response. I am currently using NAMD 2.12 and the tutorial i was following is https://www.ks.uiuc.edu/Training/Tutorials/science/channel/channel-tut.pdf.

Regards,

On Tue, Oct 16, 2018 at 8:31 PM Giacomo Fiorin <giacomo.fiorin_at_gmail.com<mailto:giacomo.fiorin_at_gmail.com>> wrote:
Please check the manual for the harmonic restraint:
https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:colvarbias_harmonic
it is "centers" (plural), not "center".

In general, you should reference which version of NAMD you are using (different versions include different version of the Colvars module, of course), and the following if needed:
https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:colvars_config_changes

Lastly, please mention also what tutorial you are referring to (with link), because not all tutorials are up to date.

Giacomo

On Tue, Oct 16, 2018 at 8:09 PM Udaya Dahal <dahal.udaya_at_gmail.com<mailto:dahal.udaya_at_gmail.com>> wrote:
Hi Everyone,

I am trying to calculate PMF for a peptide chain bound to surface. I have a fixed surface and the reaction coordinate is along z-direction. For umbrella sampling, I need to restrain my peptide's center of mass at specific point along z direction. For this I used one (fixed atom) from surface as a reference atom and peptides atoms as group of atoms for "distanceZ" module in namd, The script is as follows similar to that in the tutorial. The following description doesn't seem to define center of mass of the atoms for the "main" which led to crashing the simulation( my assumption). The "distance" module seems defines two center of mass groups but there is no option for restraining along particular axis. I would appreciate if there is any suggestion on how to proceed from here.
colvar {
   name Translocation
   width 0.1
   lowerboundary 40.0
   upperboundary 100.0
   lowerwallconstant 100.0
   upperwallconstant 100.0
   distanceZ {
      main {
         atomnumbers { 7301 7302 7303 7304 7305 7306 7307 7308 7309 7310 7311 7312 7313 7314 7315 7316 7317 7318 7319 7320 7321 7322 7323 7324 7325 7326 7327 7328 7329 7330 7331 7332 7333 7334 7335 7336 7337 7338 7339 7340 7341 7342 7343 7344 7345 7346 7347 7348 7349 7350 7351 7352 7353 7354 7355 7356 7357 7358 7359 7360 7361 7362 7363 7364 7365 7366 7367 7368 7369 7370 7371 7372 7373 7374 7375 7376 7377 7378 7379 7380 7381 7382 7383 7384 7385 7386 7387 7388 7389 7390 7391 7392 7393 7394 7395 7396 7397 7398 7399
         }
      }
      ref {
         atomnumbers { 1910
         }
      }
      axis ( 0.0, 0.0, 1.0 )
   }
}

harmonic {
        name Z
colvars Translocation
        CENTER 11.5
        forceConstant 0.025
}

Regards,
Udaya


--
Giacomo Fiorin
Associate Professor of Research, Temple University, Philadelphia, PA
Contractor, National Institutes of Health, Bethesda, MD
http://goo.gl/Q3TBQU
https://github.com/giacomofiorin



This archive was generated by hypermail 2.1.6 : Tue Dec 31 2019 - 23:20:15 CST