[gmx-users] Fwd: cut in gmx clustsize

Alex alexanderwien2k at gmail.com
Sat Oct 5 17:40:15 CEST 2019


Any comments, please?

Thanks,
Alex

---------- Forwarded message ---------
From: Alex <alexanderwien2k at gmail.com>
Date: Fri, Oct 4, 2019 at 13:58
Subject: cut in gmx clustsize
To: <gmx-users at gromacs.org>


Dear all,
In clustering what is usually defined as a criteria is a cut-off so that if
two atoms or molecules are closer to each other less than the cut-off then
those two are being counted to be in a cluster, I see even that in the
source code of gmx clustsize where
...
/* If distance less than cut-off */
if (bSame)
{
/* Merge clusters: check for all atoms whether they are in
* cluster cj and if so, put them in ci
*/
for (k = 0; (k < nindex); k++)
...
However, what is puzzling me is the definition of the -cut as "Largest
distance (nm) to be considered in a cluster" in gmx clustsize man page.
So, would you please elaborate that further?

Also, how the default cut-off of 0.35 nm has been indicated?

And finally, in the case of using -mol option to consider the molecules
rather than atoms in clustering, I wonder if the cut-off is the distance
between center of mass of the molecule?

Regards,
Alex


More information about the gromacs.org_gmx-users mailing list