normal Energy minimization of PW fails

  • denner
  • denner's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 10 months ago #5594 by denner
Energy minimization of PW fails was created by denner
Hi everyone,

I am trying to set up a system with PW, but something strange keeps happening during the energy minimization. I am pretty sure that I am missing something obvious, but I cannot figure out what. The same problem also arises when I just use PW, so something fundamental must be wrong...

If I take the PW .gro file from the MARTINI website and run an energy minimization on that, the molecules align themselves while stretching. The minimization runs for a little while with the energy and Fmax decreasing but then Fmax suddenly increases sharply. Looking at the trajectory one sees that a WM of one PW and a WP of another "goes together".

My .mdp file is:
integrator               = steep
dt                       = 0.04  
nsteps                   = 5000
nstcomm                  = 100

nstlist                  = 10
ns_type                  = grid
pbc                      = xyz
rlist                    = 1.4

coulombtype              = Shift
rcoulomb_switch          = 0.0
rcoulomb                 = 1.2
epsilon_r                = 2.5 ;
vdw_type                 = Shift
rvdw_switch              = 0.9
rvdw                     = 1.2

constraints              = none 
constraint_algorithm     = Lincs
continuation             = no
lincs_order              = 4
lincs_warnangle          = 90

I have all the files here:
www.dropbox.com/s/6fam248rbpingzx/failed_em.tar?dl=1
If you look at the trajectory in VMD, look at index 694.

I would appreciate any help/comments :-)

Please Log in or Create an account to join the conversation.

  • denner
  • denner's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 8 months ago #5749 by denner
Replied by denner on topic Energy minimization of PW fails
Just bumping the thread, hoping someone has some suggestions :-)

Please Log in or Create an account to join the conversation.

More
7 years 8 months ago #5750 by mnmelo
Replied by mnmelo on topic Energy minimization of PW fails
Hi Denner,

Sorry for the long time to reply. Good that you bumped this.

One thing to try is to use stiff bonds instead of constraints when minimizing. If you look at the .itp part of the polarizable water you'll see there's a whole section of constraints you can replace by bonds.

Try that, but **for minimization purposes only**. Do change it back to constraints for MD runs.

Good luck and let me know if this solves the issue,
Manel

Please Log in or Create an account to join the conversation.

  • denner
  • denner's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 8 months ago #5758 by denner
Replied by denner on topic Energy minimization of PW fails
Thanks for the response.

I am aware of this and normally also use these bonds for minimizations, which I have also done here.

Please Log in or Create an account to join the conversation.

More
7 years 8 months ago #5759 by mnmelo
Replied by mnmelo on topic Energy minimization of PW fails
Hmm, this is a tricky one.

I can reproduce your results. I have encountered similar problems in the past when the box is at a pressure higher than the equilibrium pressure. In those cases waters sit at short distances from one another, which makes it easier for the opposite charge attraction to trump the LJ repulsion of the water centers. Perhaps in this particular waterbox from the website two water molecules are for some reason quite close, and the same happens.

In any case, minimizing with constraints instead of bonds seems to do the trick (exactly the opposite of what I first suggested!).

I'll try to update the website with a waterbox that does minimize properly with the recommended method. Thanks for reporting this!

Manel

Please Log in or Create an account to join the conversation.

  • denner
  • denner's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 8 months ago #5760 by denner
Replied by denner on topic Energy minimization of PW fails
Thanks again :-)

I agree that for this particular water box, it works fine to use constraints instead.
However, I get too many LINCS warnings when trying to set up my other system (which is why I switched to stiff bonds to begin with): is that a sign of a "broken" starting structure or should I try to use your new waterbox, to solvate my system, once you have updated the website?

Please Log in or Create an account to join the conversation.

More
7 years 8 months ago #5762 by mnmelo
Replied by mnmelo on topic Energy minimization of PW fails
Ah, nailed the cause, and found a solution.

Apparently the stiffness of 50000 kJ/mol/nm^2 in the .itp is too soft for minimization when waters are already at intermolecular distances close to equilibrium. I guess the water-dummy_charge bond just stretches a bit and overlaps with other charges occur.

Setting the stiffness of the bonds to 100000, for minimization purposes, did the trick for me.

Please Log in or Create an account to join the conversation.

  • denner
  • denner's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
7 years 8 months ago #5763 by denner
Replied by denner on topic Energy minimization of PW fails
Thank you so very much!
Should have tried that earlier :-)

Please Log in or Create an account to join the conversation.

Time to create page: 0.108 seconds