normal NVT & NPT run for coarse grained simulation

  • xudongh
  • xudongh's Avatar Topic Author
  • Visitor
11 years 4 months ago - 11 years 4 months ago #1231 by xudongh
Hello all,

I'm trying to run a NVT and NPT equilibration on this CG model following the Protein In Water tutorial. Is there a supplied NVT.mdp or NPT.mdp file for this type of CG system? I adapted another source's nvt and npt.mdp file, and I tried to run it, i get the following error:

Fatal error:
There is no domain decomposition for 4 nodes that is compatible with the given box and a minimum cell size of 5.2375 nm
Change the number of nodes or mdrun option -rcon or -dds or your LINCS settings
Look in the log file for details on the domain decomposition
For more information and tips for troubleshooting, please check the GROMACS
website at www.gromacs.org/Documentation/Errors

I don't think I can change the node because that was decided way earlier on in the run many steps before, and I would imagine if I change the node, i change the entire structural property.

Any suggestions? Let me know.

**Is it because we coarse grained it, which leaves it not enough atoms?

Thanks :)

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

More
11 years 4 months ago #1234 by xavier
Replied by xavier on topic NVT & NPT run for coarse grained simulation
Ok this happens with small systems ... the code can not decompose the system in four parts in a dynamic way ...

try to add the "-pd" (particle decomposition) option to mdrun ... should fix it

xudongh wrote: Hello all,

I'm trying to run a NVT and NPT equilibration on this CG model following the Protein In Water tutorial. Is there a supplied NVT.mdp or NPT.mdp file for this type of CG system? I adapted another source's nvt and npt.mdp file, and I tried to run it, i get the following error:

Fatal error:
There is no domain decomposition for 4 nodes that is compatible with the given box and a minimum cell size of 5.2375 nm
Change the number of nodes or mdrun option -rcon or -dds or your LINCS settings
Look in the log file for details on the domain decomposition
For more information and tips for troubleshooting, please check the GROMACS
website at www.gromacs.org/Documentation/Errors

I don't think I can change the node because that was decided way earlier on in the run many steps before, and I would imagine if I change the node, i change the entire structural property.

Any suggestions? Let me know.

**Is it because we coarse grained it, which leaves it not enough atoms?

Thanks :)

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

  • xudongh
  • xudongh's Avatar Topic Author
  • Visitor
11 years 4 months ago #1235 by xudongh
Replied by xudongh on topic NVT & NPT run for coarse grained simulation
@ xavier,

When i try -pd in my mdrun, the run doesn't last for 5000 steps, infact it ends after step 0. Is it because the system is too small for a 5000 step run?

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

More
11 years 4 months ago #1236 by xavier
Replied by xavier on topic NVT & NPT run for coarse grained simulation
No, then you have another problem! Try to identify it!

xudongh wrote: @ xavier,

When i try -pd in my mdrun, the run doesn't last for 5000 steps, infact it ends after step 0. Is it because the system is too small for a 5000 step run?

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

  • xudongh
  • xudongh's Avatar Topic Author
  • Visitor
11 years 4 months ago #1237 by xudongh
Replied by xudongh on topic NVT & NPT run for coarse grained simulation

xavier wrote: No, then you have another problem! Try to identify it!

xudongh wrote: @ xavier,

When i try -pd in my mdrun, the run doesn't last for 5000 steps, infact it ends after step 0. Is it because the system is too small for a 5000 step run?


Oh crap. That reminds me when I ran the em.mdp (energy minimization of the solvated structure) it was also set to 5000 steps, it did run all 5000 steps, but the run was rather quick compared to all atom based system. Was that a hallmark feature of coarse grained the structure so it cuts down computational time or did I mess up there too?

Thanks,

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

More
11 years 4 months ago #1238 by xavier
Replied by xavier on topic NVT & NPT run for coarse grained simulation
!? Yes CG simulations cut down in computational time ...

xudongh wrote:

xavier wrote: No, then you have another problem! Try to identify it!

xudongh wrote: @ xavier,

When i try -pd in my mdrun, the run doesn't last for 5000 steps, infact it ends after step 0. Is it because the system is too small for a 5000 step run?


Oh crap. That reminds me when I ran the em.mdp (energy minimization of the solvated structure) it was also set to 5000 steps, it did run all 5000 steps, but the run was rather quick compared to all atom based system. Was that a hallmark feature of coarse grained the structure so it cuts down computational time or did I mess up there too?

Thanks,

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

  • xudongh
  • xudongh's Avatar Topic Author
  • Visitor
11 years 4 months ago #1243 by xudongh
Replied by xudongh on topic NVT & NPT run for coarse grained simulation

xavier wrote: !? Yes CG simulations cut down in computational time ...

xudongh wrote:

xavier wrote: No, then you have another problem! Try to identify it!

xudongh wrote: @ xavier,

When i try -pd in my mdrun, the run doesn't last for 5000 steps, infact it ends after step 0. Is it because the system is too small for a 5000 step run?


Oh crap. That reminds me when I ran the em.mdp (energy minimization of the solvated structure) it was also set to 5000 steps, it did run all 5000 steps, but the run was rather quick compared to all atom based system. Was that a hallmark feature of coarse grained the structure so it cuts down computational time or did I mess up there too?

Thanks,



@Xavier.

For a MD.mdp run, is 5 million step on a cg system too much and redundant?

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

More
11 years 4 months ago #1249 by djurre
Replied by djurre on topic NVT & NPT run for coarse grained simulation
That depends on what you want to do/investigate. So try it out and see what happens during the simulation and if you can answer your scientific questions with the data you get.

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

More
5 years 4 months ago #7902 by riccardo
Replied by riccardo on topic NVT & NPT run for coarse grained simulation
See here:
www.gromacs.org/Documentation/Errors#The...ze_or_decrease_rlist

Something is going very wrong - aka, your system is exploding or similar. Note that you cannot decrease rlist at will. This is a parameter which goes into the parametrization of a force field, thus it is usually fixed to the value which has been used to parametrize the force field. If you change it, who knows in which way you'll change the behavior of the force field.

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

More
5 years 4 months ago #7910 by riccardo
Replied by riccardo on topic NVT & NPT run for coarse grained simulation
You usually do not want/need to energy minimize your system after an NPT. One usually energy minimizes as a first step (to remove high forces due to too close contact between particles in the system), then equilibrate the temperature (NVT) and subsequently the density/pressure (NPT) - or you can do both at once going straight to NPT (usually fine with Martini). Like I wrote in my previous post, there must be something going very wrong in your system. Have you taken a look at the gro/pdb file at the end of the minimization? Do things look alright (homogeneous density of particle, no bad overlaps, etc.)? How big is your box?

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

Time to create page: 0.095 seconds