Re: [AMBER] kclust help

From: Daniel Roe <>
Date: Thu, 2 Jul 2015 11:08:11 -0600


I don't think kclust is maintained by any Amber developers, so I'm not sure
this is the best place to ask for help with it (although maybe some users
have experience with it?).

What is the composition of your system and what are you trying to cluster
on (i.e. what is your distance metric)? Generally with trajectories
generated with iwrap=1 and multiple solute molecules of interest the
recommended advice is to use 'autoimage' before processing.

You could also try using one of the clustering algorithms implemented in


On Thursday, July 2, 2015, Harmeet Kaur <> wrote:

> Dear Amber Users,
> I have used the kclust programme for clustering my trajectory (from 3
> different runs of 3 different PDBs, totalling 3000 structures). When radius
> is set to 5, the no. of clusters obtained is 78 (for 3000 structures). I
> have used iwrap=1 in my production run. While generating the dry
> trajectory, I tried using 'unwrap' but that gave me even higher no. of
> clusters (close to 1000). Then I tried using 'image center familiar' and
> got 78 clusters which is still large for analyses.Kindly suggest what
> should be done to minimize the no. of clusters without increasing the value
> of radius. I'd also be grateful if anyone can send a reference on
> trajectory analysis by kclust since I want to know the maximum value that
> can be set for radius during clustering.
> Thanks in advance
> Regards
> Harmeet
> _______________________________________________
> AMBER mailing list
> <javascript:;>

Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 307
Salt Lake City, UT 84112-5820
(801) 587-9652
(801) 585-6208 (Fax)
AMBER mailing list
Received on Thu Jul 02 2015 - 10:30:02 PDT
Custom Search