[gmx-users] g_cluster for ligand clustering

[mbalint at caesar.elte.hu]

Dear all,

I have a trajectory file of a protein+ligand binding simulation.
What I would like to do is to cluster the ligands positions from the  
simulation trajectory. But although I have tried quite a few  
possibilities, the RMSD calculation using g_cluster will change the  
reference coordinates with respect to the receptor. I have tried to  
select the ligand's group for the RMSD calculation, but this option  
gave me only one cluster (which cannot be real), even if I have  
decreased the cutoff.

Command line used:
g_cluster -f full_4.xtc -s pr.tpr -method gromos -cutoff 0.1 -cl proba.pdb

I have also tried to make a clustering from a cumulative pdb file  
(containing all the frames) with -s being the reference pdb file, but  
the result was the same.
I assume there must be a relatively easy way to do this task, but now  
I just can't see how.

Could you please let me give me another idea how can I cluster my ligand?

With the kindest regards,
Monika Balint
PhD Student