[gmx-users] distance options for g_cluster

[Mark Abraham]

OZGE ENGIN wrote:
> Hi all,
> 
> I am trying to understand what type of distance can be used with g_cluster? 
> 
> In the manual, it is written that  RMS deviation after fitting or RMS deviation of atom-pair distances can be used to define the distance between structures in the first paragraph. 
> However, in the "other options" part, two options are indicated for 
> -dista option, which indicates the type of distance that will be used with  g_cluster. They are RMSD of distances or RMS deviation. Consequently, the information in the "first paragraph" does not coincide with that given in the "other options" part for -dista option. 
> 
> Maybe I could not understand. Could anyone explain this in a clearer way, please?

I agree that it is not very clear what is meant in the documentation, 
and I have wondered myself what it means. This is one of the reasons 
that the source code is made available.

I had a quick look, and "-dista yes" looks like it implements the "RMS 
deviation of atom-pair distances" quoted above. That is, it forms an 
atom-pair distance matrix for each structure in the trajectory, and then 
does an all-against-all root-mean-square comparison of those matrices to 
generate the structure comparison matrix for clustering. You'd be well 
advised to verify that yourself. It's all in src/tools/gmx_cluster.c

Mark