[gmx-users] simulation problem whit lincs

[Anton Feenstra]

Osmany Guirola Cruz wrote:

> I am doing a simulation of system with one protein and a peptide
> i do all steps necesary for the simulation(minimization, genion
> ,position restrain etc...) when i run the final step(MD) runs whitout
> problem during some time.. these is the error
> Step 161731, time 323.462 (ps)  LINCS WARNING

This can be simply bad luck: every once in while two atoms will collide
a bit to fast, rebound and escalate. My workaround for this is to take
the last .trr frame with positions *and* velocities you have, and restart
from there, but with a minimal change by introducing slight round-off
errors by a conversion to and from .gro file (with 6 decimals), like this:

#> trjconv -f traj.trr -dumpo 300 -o frame_300ps.gro -ndec 6
#> trjconv -f frame_300ps.gro -s topol.tpr -o frame_300ps.trr
#> tpbconv -f frame_300ps.trr -s topol.tpr -o restart.tpr

Then, run mdrun -f restart.tpr. If the problem re-occurs, you have a
problem in your conformations and should check what might be the cause.
If it runs on happily, there is (IMHO), not a problem after all!


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/       |
|             | "If You See Me Getting High, Knock Me Down"           |
|             | (Red Hot Chili Peppers)                               |
|_____________|_______________________________________________________|