Hello lammps user,
I am running a simulation consists of two protein by implicit solvent model. When I run the simulation, there is an error appear as below:
Bond/angle/dihedral extent >half of periodic box length.
I tried to fix the problem by changing the box size which was not working. On the other hand, we you minimize the system, simulation runs but protein structure splits
I would be thankful if you help me out with this.
Regards
Hello lammps user,
I am running a simulation consists of two protein by implicit solvent model. When I run the simulation, there is an error appear as below:
Bond/angle/dihedral extent >half of periodic box length.
does this happen right at the beginning or later during the run?
I tried to fix the problem by changing the box size which was not working. On the other hand, we you minimize the system, simulation runs but protein structure splits
the atoms go where the potentials tell them to. if that is not what you expect then either your topology data in incorrect, or your geometry not consistent with the topology or your force field parameters incorrect or not suitable for the kind of system you have.
I would be thankful if you help me out with this.
it is impossible to provide any specific advice on “my simulation doesn’t work as expected”
It happens at the beginning
At the beginning, I replicate my box in y direction then this error occurs. I even run the simulation for just one protein and increased the box size. This error appeared when I increased the box size too.
Please let me know if sending any part of input or data file can help you with this problem.
if it happens right from the beginning, than most likely your data file has problems, e.g. bonds between atoms far apart or inconsistent image flags or unsuitable box parameters or a combination of those.
Thank you for your comments Alex.
Can you please help me how I can check this inconsistency and overcome it? I literally have no idea how to start checking it.
it all depends on how you prepare the data file.
I would start by not building an input for your system right away, but with something much simpler and smaller.
Then you can manually verify the data file yourself. also that teaches you how to “read” a data file.
With that example you can then generate an input and then run for zero steps and write the configuration as a dump file and compare.
If you get a warning, you can address it properly. You can effectively build your entire workflow with this toy system.
If that all works out and all issues are addressed, you can build a more complex input and do the same thing again. It will be easier to spot inconsistencies since you are more familiar with the file format and have addressed issues in the workflow itself.
and so on and so forth.
this is really just common sense.