AMBER Archive (2009)
Subject: Re: [AMBER] ptraj usage for clustering of protein ensembles manual or tutorial
From: Andrew Voronkov (drugdesign_at_yandex.ru)
Date: Fri Dec 04 2009 - 13:09:58 CST
I actually want to divide trajectory of 5000 snapshots from 5 nanoseconds into 10-20 representative structures as it is done in relaxed complex scheme, but in my case i dont have a ligand, just want to simulate the protein flexibility by usage of most common structures with rigid docking. So what procedure is best for such a purpose and is there anywhere more detailed description of these procedures?
04.12.09, 10:01, "Thomas Cheatham III" <tec3_at_utah.edu>:
> > trajin fzd2_w_md5.mdcrd
> > cluster out avg-1 representative none average none all none averagelinkage clusters 1 mass rms
> > "(:1-240) & @CA,C,N,O"
> > go
> The "none" keyword means you are choosing not to dump any coordinate
> output for the "representative" snapshot (that closest to the centroid),
> the "average" over each cluster, and dumping of "all" the frames in each
> Also, I do not not understand why you would want to cluster into a single
> (cluster 1) cluster as this will simply give back your original
> cluster out c-out representative pdb average pdb averagelinkage \
> clusters 10 mass rms :1-240_at_CA,C,N,O
> will create
> c-out.avg.n <- average structures for each cluster as PDB
> c-out.rep.n <- representative " "
> c-out.n <- trajectories in mdcrd format for each cluster
> -- tec3
> AMBER mailing list
AMBER mailing list