CLUST'EM-D README FILE
Clust'em-D is a hierarchical clustering tool specially designed to analyze molecular
dynamics trajectories i.e. a set of conformations or frameworks resulting from a time-
resolved molecular simulation.
Required packages:
Clust'em-D requires the installation of the following packages for python:
- numpy: available at http://numpy.scipy.org/
- scipy: available at http://numpy.scipy.org/
- matplotlib: retrieved from http://matplotlib.sourceforge.net/
- Bio.PDB: available at http://biopython.org/DIST/docs/api/Bio.PDB-module.html.
- rpy2: downloaded from http://rpy.sourceforge.net/rpy2.html (Requires R package to be installed(http://www.r-project.org/))
Clust'em-D also requires some libraries for R:
- Bio3D: available at http://bio3d.pbworks.com/w/page/7824467/FrontPage
- gclus: available at http://cran.r-project.org/web/packages/
Clust'em-D commands:
The simplest Clust'em-D run command follows the pattern:
python clustemd.py -i [input] -o [output] -f [format]
Where:
[-h, --help] for help
[-i, --infile] directory path where input files are located
[-o, --output] the name of the directory for the resulting files
[-f, --format] format used for the analysis
"dcd" when a dcd and a pdb files of a trajectory are included
"pdb" each pdb file in the "-i" specified directory corresponds to a single snapshot of a MD trajectory
[-c, --cutoff] the number of levels considered in the cluster tree
by default the number of levels is 3
[-l, --list] list with the proper order of pdb files
by default, the program will follow the bash "ls" command order
Clust'em-D output:
In the output directory, apart from a plain text report (report.out), the user will be provided with a .tex file that may
be compiled with, for instance, pdflatex to straightforwardly generate a step-by-step report including the plots
created by Clust'em-D.