On Thu, Apr 26, 2018 at 11:10 AM, Alexander Alexander <[email protected]> wrote: > Hi, > > On Thursday, April 26, 2018 at 12:23:36 PM UTC-4, Christoph Junghans wrote: >> >> On Thu, Apr 26, 2018 at 9:12 AM, Alexander Alexander >> <[email protected]> wrote: >> > Thank Christoph. >> > >> > On Thursday, April 26, 2018 at 10:49:25 AM UTC-4, Christoph Junghans >> > wrote: >> >> >> >> On Thu, Apr 26, 2018 at 1:37 AM, Alexander Alexander >> >> <[email protected]> wrote: >> >> > >> >> > Dear all, >> >> > Using IBI over nonbonded parameters I intend to parameterize the CG >> >> > parameter for a molecule which has 3 beads in 2 types. The initial >> >> > CG-configuration.gro was extracted from the trajectory of an >> >> > atomistic >> >> > simulation over the system which I am pretty sure that it has been >> >> > enough >> >> > equilibrated with different ensemble and for enough computational >> >> > time >> >> > as >> >> > all of its properties are being stable by time. Even in each CG step >> >> > I >> >> > do >> >> > the minimization (pre.simulation). However, the CG nonbonded >> >> > parametrization >> >> > crashes due to the issue of "1 of the 692 bonded interactions could >> >> > not >> >> > be >> >> > calculated because some atoms involved moved further apart than the >> >> > multi-body ...". >> >> Usually that error means your system is exploding. >> >> Have a look at the simulation movie with vmd. >> > >> > Yes, I see how the system somehow collapses. BUT: Actually it works fine >> > (it >> > is now in step_003) with "mpirun -np 1 gmx_mpi mdrun" and no explosion >> > happens in the system, which means the system itself is enough stable, >> > but I >> > really need more than 1 slot and more than 1 slot even with -np 2 the >> > crashing shows up again. So, I think it is more the matter of >> > computation >> > rather than the physics of the system, am I right? I want -np 16. >> On a single node the domains are just bigger and hence it takes longer >> to trigger the " bonded interactions could not be calculated" > > Each step now run for 1.5ns and it works. > You mean even if I get converge result using 1 slot and 1.5ns in each step, > the potential might be wrong? I was saying, that just because it isn't crashing on one core, your results still don't need to be right. On one core, seeing the "bonded interactions could not be calculated" might just take longer.
>> >> >> >> >> Try to replace the related bonded interaction with a potential with a >> >> stronger spring constant. >> > >> > I have already changed the following lines one by one or together I hope >> > correct!?; >> > [ bonds ] >> > 1 2 8 1 1 ; 1:bond-LA-C:1 ------ changed -----> 1 2 X Y Z with >> > different combinations of X =1 or2, Y=1,4 and Z = 1000 ---6000 >> > [ angles ] >> > 1 2 3 8 1 1 ; 1:angle-LA-C-RA:1 ------ changed -----> 1 2 3 X Y' >> > Z >> I think that is correct, but I would consult the gromacs manual to be >> sure. >> With the different combination I used the job still crashes. Then it sounds like that something is wrong with your topology... Christoph > > > Regards, > Alex >> >> Christoph >> > .... >> > But still the problem is here. >> > >> > Regards, >> > Alex >> >> >> >> >> >> Christoph >> >> > I have been trying most of the suggested solutions for such an error, >> >> > but no >> >> > success. I even reduced the number of beads from my initial mapping >> >> > (5 >> >> > beads) to the current one (3 beads), but it crashes even earlier. >> >> > The materials for this system have been shared in the below link, I >> >> > would be >> >> > so appreciated if you could kindly help me solve the problem. >> >> > https://drive.google.com/open?id=12VIJ3W9BgE0IlEls8ZlJ1uIAG3jqbNG0 >> >> > >> >> > Thank you very much. >> >> > >> >> > Best regards, >> >> > Alex >> >> > >> >> > -- >> >> > You received this message because you are subscribed to the Google >> >> > Groups >> >> > "votca" group. >> >> > To unsubscribe from this group and stop receiving emails from it, >> >> > send >> >> > an >> >> > email to [email protected]. >> >> > To post to this group, send email to [email protected]. >> >> > Visit this group at https://groups.google.com/group/votca. >> >> > For more options, visit https://groups.google.com/d/optout. >> >> >> >> >> >> >> >> -- >> >> Christoph Junghans >> >> Web: http://www.compphys.de >> > >> > -- >> > You received this message because you are subscribed to the Google >> > Groups >> > "votca" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> > an >> > email to [email protected]. >> > To post to this group, send email to [email protected]. >> > Visit this group at https://groups.google.com/group/votca. >> > For more options, visit https://groups.google.com/d/optout. >> >> >> >> -- >> Christoph Junghans >> Web: http://www.compphys.de > > -- > You received this message because you are subscribed to the Google Groups > "votca" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at https://groups.google.com/group/votca. > For more options, visit https://groups.google.com/d/optout. -- Christoph Junghans Web: http://www.compphys.de -- You received this message because you are subscribed to the Google Groups "votca" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/votca. For more options, visit https://groups.google.com/d/optout.
