Say, I have performed a NVT simulation, and have got a trajectory file. After that, I would like to extract the final frame and change the dimension of the system, and use it for another NVT simulation. I have tried to do that using the trjconv and editconf commands, however, when mdrun is performed, there is an error message, saying there is a segmentation fault (core dumped) error. How can I avoid such kind of error?
If you do not need to do an exact continuation of your NVT run, the simplest way is to manually modify final gro file and use it as an input to next-stage grompp. The periodic box dimensions are on the last line (check http://manual.gromacs.org/online/gro.html for the meaning). If mdrun did not create any gro file (due to e.g. simulation blow-up), you can extract the last frame using trjconv.
Are you making the box larger or smaller? If the box gets larger, I see no reason why this should fail. If you're making it smaller, then you will have clashes at the edges of your box that may cause instability.
I am going to make the box smaller. To be honest, I am still a beginner in using gromacs. Is there any way to avoid such clash? The system has a 3D periodic boundary condition.
Ad hoc changes to the vectors of an equilibrated system will lead to unavoidable clashes; there's no way around that. What is your objective? If you're trying to simulate at different densities, for instance, you could run an NPT simulation at higher pressure to compress the system, then run NVT once the system compresses to the desired density.
Thanks for your reply. I have solved the problem. There was no error when I reduced the volume by 15%. In fact, I am trying to prepare a well-relaxed polymer structure as an initial structure for a 60-100ns NPT simulation. This method was developed by William Goddard. Attached is the paper.
Article Hildebrand and Hansen solubility parameters from Molecular D...
- Badges
- Science topic