I want to measure the number of monomers, dimer, ... between the same molecule in my simulation box. I am using the following code:
gmx clustsize -f mdrunps.xtc -s mdrun.tpr -n index.ndx -tu ps -b 1 -e 2 -hc hist.xvg
This command considers molecules as a cluster if the largest distance between them be 0.35 nm (0.35 can be changed with "-cut" option).
But I want to consider the distance between the center of mass as criteria to form a cluster.
Any help is appreciated.
Hi Ali,
Use gmx distance specifying the atoms of interest.
gmx distance -n index.ndx -f traj.xtc -s topol.tpr -oav -oall"
Ali Ghamartale
Hi,
I don't think you can specify the COM as an option for gmx clustsize. However, I believe just specifying the index group of all molecules of interest (make sure to use full molecule, not any subgroup), should give you the desired result. Also, play with the 'cut' option as well. You can also plot the number of clusters over time with the '-nc' option in gmx clustsize like this :
Dear Martiniano Bello ,
Thank you so much for responding. I wonder if g_distance ( and/or the command that you mentioned) can calculate the distance between every pairs of 25 molecules. To see if they make cluster or not I need to know the distance between molecules 1-2, 1-3, 1-4, ..., 1-25 and 2-3, 2-4, ..., 2-25 and ... and 24-25 in each time step. I appreciate you if you have more idea
Dear Vikas Dubey
Thank you so much for your comment. To solve to COM problem, I got an Idea! I extracted the coordination of COM for a single molecule and I find the closest atom to COM (all done by VMD). Then, identified that atom in my index file. Then, as you said I change -cut. I would like to know your idea about this approach.
about '-nc' option you are right but the problem is I need to know how many clusters are monomer, how may are dimmer and so on. That's why I used '-hc' command. I wonder if you have any better Idea.
- Badges
- Science topic