my production run generated 2000 frames but when i i load my md.gro and md.trr file in vmd, vmd shows only 252 frames. what might be the problem? why all the frames cannot be seen? also, i want to report here that the md.gro file generated during my md run in gromacs shows that the protein has moved out of the water box.
please guide me as to why vmd is showing less number of frames from the 2000 generated frames.
Just look into your .mdp file and check what is the interval of saving "nstxout". What is the nsteps? if nstep/nstxout is 2000, so probably your simulation has been crashed before reaching the end step. So just check your *.edr file by g_energy code to double check that (see how many steps has been done successfully).
As Justin mentioned, this is not a problem and it happens during saving coordinates. To get rid of this you can do this:
trjconv -f *.trr -s *.tpr -pbc atom to put all atoms in the box again
trjconv -f *.trr -s *.tpr -pbc whole to make the protein molecule whole again (This should be run on the output of the previous trjconv run. you may need to run this couple of times depending on the size of your macro molecule.)
I hope this helps
VMD only loads what it finds. If you have fewer frames than you thought, then either (1) you didn't save at the interval you thought or (2) the run stopped prematurely. What does gmxcheck tell you about the .trr file? Does the .log file from the run indicate that the run stopped unexpectedly?
There is also no such thing as "outside" of a periodic box. That's the point - the system is infinite :) It's totally normal for these visualization artifacts to arise, and they are quickly fixed with the proper call to trjconv to center the protein and/or re-wrap the unit cell.
Just look into your .mdp file and check what is the interval of saving "nstxout". What is the nsteps? if nstep/nstxout is 2000, so probably your simulation has been crashed before reaching the end step. So just check your *.edr file by g_energy code to double check that (see how many steps has been done successfully).
As Justin mentioned, this is not a problem and it happens during saving coordinates. To get rid of this you can do this:
trjconv -f *.trr -s *.tpr -pbc atom to put all atoms in the box again
trjconv -f *.trr -s *.tpr -pbc whole to make the protein molecule whole again (This should be run on the output of the previous trjconv run. you may need to run this couple of times depending on the size of your macro molecule.)
I hope this helps
Agree with Abolfazl Noorjahan. Crashed simulation seems the only reason.
the .log file also says that the run is complete and all 2000000 steps have been run and the checkpoint file has also been written every 15 minutes throghout the run till 2000000 steps. (according to the .log file)
With a system of 355397 atoms, VMD is probably just running out of memory to load all of the frames. Clearly all your data are intact, though. Strip the water and anything else unnecessary using trjconv and try again.
If VMD is running out of memory, it just crashes (That's the condition in Mac). I think you may try to set a STRIDE as 100. If VMD can load 20 frame. That means your trajectory file is OK. You need to try other command-line tools like AMBERTOOLS to deal with it. Or you can divide the big trajectory file to several smaller fragments with AMBERTOOLS or other tools. If you can, I recommend you to convert your trajectory file to DCD. It's easy-going for VMD.
Load the xtc file, I'm not sure of how VMD is able to handle properly the trr file, so it may mis-behave because atom records are now read properly.
thankyou all for your kind help....u all are correct....vmd was running out of memory...it is working now :) ...thankyou again
Hi guys!
I 'd had the same problem. I've followed your advice and removed water. Although it has worked and improved frame visualization but VMD still stops before reading the last part. It seems that .xtc file is OK.
Should I remove some frames from the beginning? How?
Remove water molecules as that will definitely reduce the memory load.
Of course I did. Please consider I set nstxtcout for 5000 frame and gro file contains about 6700 atoms. VMD stops reading around frame of 4002.
Hi Homa,
I am experiencing the same problem (with another software than VMD), and I want to split the trajectory into two (or more) smaller trajectories. If you solved your problem by splitting the trajectory, can you please share how did you split it?
We encountered the problem of incomplete loading in VMD in Windows. I believe it is a problem of memory allocation for VMD under Windows. When we run VMD under LInux, we had no such a problem even for the same RAM available in the PC
Hi all i have a question i have completed 100 ns steps by using Gromacs. i want to ask you if i want to visualize it through vmd so there is need to join all .gro files like all .xtc files ? if yes then how because if i use trjcat -f *.gro -o traj-out-all.gro
this command i am facing problem with an error
Fatal error:
Can not write a gro file without atom names
How to deal with it? please guide me.
You can concatenate .gro file and VMD will display them, but this is the most inefficient way to do it in terms of memory. A multi-frame .gro file (being a plain-text format) is much larger in terms of disk space than a normal trajectory (compressed .xtc). I recommend you work with normal trajectory formats. Also, by default, GROMACS only writes the final frame in .gro format, so you're missing nearly all of the information from the simulation if you only visualize those.
Justin Lemkul thank you for your suggestion. it is really helpful to me.
- Badges
- Science topic