I can easily do it by using trjconv -b and -e option if I need to extract only one snapshot. However, if have a list of suppose 5000 snapshots, not in particular order, it gets very time-consuming. Any suggestions will be highly appreciated.
I am doing as follows:
for i in `cat index_file`; do echo -e "System"|gmx trjconv -f $TRR -s $TPR -b $i -e $i -o ${i}.trr; done
However, it takes a lot of time as every time it starts from the beginning then reaches to the specified frame in the index_file. However, I wanted something that it continues reading from the previous frame so that saves the time.
If you just need to pull out select frames based on individual time steps (a long list of integers), you can feed the 'trjconv' command of GROMACS an index file:
trjconv -s conf.tpr/gro -f traj.xtc/trr -o frames.xtc -fr frames.ndx
The important file is frames.ndx, of which I give an example here:
`[frames]
245621
98664
313864
35
`
The way I wrote it, this code generates another trajectory, but one that specifically contains the frames you want.
From this you can use VMD or GROMACS to extract individual .pdb files.
Using the bash script.
1. Create a list in a script, and visit the element using loop.
2. Perform the trjconv command
3. End of the loop
Dear Liang,
Thanks for the suggestion. I am doing as follows:
for i in `cat index_file`; do echo -e "System"|gmx trjconv -f $TRR -s $TPR -b $i -e $i -o ${i}.trr; done
However, it takes a lot of time as every time it starts from the beginning then reaches to the specified frame in the index_file. However, I wanted something that it continues reading from the previous frame so that saves the time. Is that what you meant? Thanks.
Dear Narendra, your reply remind me that this method may takes a lot of time. As far as i know, several frames which are not discontinueous may not be extracted from the trajectory file using the trjconv one time.
Another method you may try
1. Loading your trajectory file into the VMD software 2. TCL Commands, such as mol (details see Chapter "Tcl Text Interface" in the VMD manual) can help you to get the information of a certain frame. Using the list in the tcl script in VMD, you can extract the frames you wanted by scanning your trajectory files one time.
This link will be helpful...
http://www.ks.uiuc.edu/Training/Tutorials/vmd/tutorial-html/node3.html
If you just need to pull out select frames based on individual time steps (a long list of integers), you can feed the 'trjconv' command of GROMACS an index file:
trjconv -s conf.tpr/gro -f traj.xtc/trr -o frames.xtc -fr frames.ndx
The important file is frames.ndx, of which I give an example here:
`[frames]
245621
98664
313864
35
`
The way I wrote it, this code generates another trajectory, but one that specifically contains the frames you want.
From this you can use VMD or GROMACS to extract individual .pdb files.
Sorry Sajjad, I did not see the question until today.
I believe the snapshots are 'frame' numbers, not time-steps.
I have not tested this with an .xtc containing time-steps not initialized to zero, so I am not certain, but trjconv should report its own results.
I am not familiar with what 'sub' does to the output, but I also put all of my frames in a single column - maybe it makes no difference, but it is something you might test.
Dear Dr. Gáspári,
We also run into a situation where we want to restrain all the backbone dihedral angles in a protein molecule to their starting crytallographic values. It's nice to know that you have a script that can do it in Gromacs in an automatic way. I am wondering if you can share your scripts with us? Thanks.
Thanks to Liang Zhao, Narendra Kumar and all others who posted to this thread. This was an excellent and very helpful discussion of the various ways one might extract selected frames from a trajectory. For what it is worth from a relative beginner with GROMACS I'd suggest Anu's and Jacob's answers are the easiest to implement.
- Badges
- Science topic