[AMBER] DBscan gives me errors in clustering

From: Debarati DasGupta <debarati_dasgupta.hotmail.com>
Date: Wed, 11 Mar 2020 15:40:34 +0000

Hi I am trying to cluster my trajectory for a simple protein in TIP3P water simulation


My input for dbscan clustering is
parm strip_2XXX.prmtop
trajin 2XXX_stripped.mdcrd 1 last
cluster D1 dbscan epsilon 2.0 kdist 3 kfile kdist_data sievetoframe rms :2-151.C,N,O,CA,CB&!.H= random out cnumvtime-dbscan.dat summary summary-dbscan.dat info info-dbscan.dat cpopvtime cpopvtime-dbscan.agr normframe repout rep-dbscan repfmt pdb singlerepout singlerep-dbscan.nc singlerep fmt netcdf avgout avg-dbscan avgfmt pdb
run


Any advice on how to troubleshoot this??




Error I get is
    CLUSTER: Using coords dataset _DEFAULTCRD_, clustering using RMSD (mask [:2-151.C,N,O,CA,CB&!.H=]) best-fit
        DBSCAN:
                Only calculating Kdist graph for K=3
                Kdist file prefix: kdist_data/
        Cluster # vs time will be written to cnumvtime-dbscan.dat
        Cluster pop vs time will be written to cpopvtime-dbscan.agr (normalized by frame)
        Pairwise distance data set is 'DB1[PWD]'
        Cluster information will be written to info-dbscan.dat
        Summary of cluster results will be written to summary-dbscan.dat
        Representative frames will be chosen by lowest cumulative distance to all other frames.
        Cluster representatives will be written to 1 traj (singlerep-dbscan.nc), format Amber NetCDF
        Cluster representatives will be written to separate trajectories,
                prefix (rep-dbscan), format PDB
        Average structures for clusters will be written to avg-dbscan, format PDB
Error: [cluster] Not all arguments handled: [ minpoints 10 epsilon 2.0 sievetoframe ]
        1 errors encountered reading input.
Warning: Set 'DB1' contains no data.
Warning: File 'cnumvtime-dbscan.dat' has no sets containing data.
Warning: File 'cpopvtime-dbscan.agr' has no sets containing data.
TIME: Total execution time: 0.1114 seconds.

Error: Error(s) occurred during execution.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Mar 11 2020 - 09:00:04 PDT
Custom Search