Let us assume that you have put the jar file or the uncompressed tar file to the directory
/home/emlurpc01/ulla/appl/treewiz In order to set up and run the TreeWiz application you have to
Set the Java CLASSPATH environment variable to point to either the jar file or to the TreeWiz base directory if you prefer to extract all the files from one of the archive files. In the latter case set the CLASSPATH to the place where the subdirectory with the name org can be found. Example (using tcsh):
With jar file: setenv CLASSPATH /home/emlurpc01/ulla/appl/treewiz/TreeWiz.jar:$CLASSPATH
With uncompressed and untarred tar file: setenv CLASSPATH /home/emlurpc01/ulla/appl/treewiz:$CLASSPATH
Modify the Properties file according to your local settings and copy it to the directory from which you want to start TreeWiz. Don't worry if some of the databases mentioned there do not exist on
your system. Start by changing the
IMAGE_DIRECTORY to /home/emlurpc01/ulla/appl/java/TreeWiz/org/eml/trees/images
and the
TREE_DIRECTORY to the directory where you have placed the tree files
in phylip format.
Run TreeWiz using the following command: java -mxXXm org.eml.trees.ForestOverview The XX has to be replaced by the number of Megabytes the Java virtual machine will be allowed to use. For trees up to several thousand nodes XX=64 should be sufficient. For the SYSTERS tree you will have to use XX=400 on a 32Bit processor machine. 64Bit processor machines like Sun or SGI workstations will need twice as much and might produce a stack overflow anyway.
Choose "Load tree file" from the menu and load a tree from the tree directory.
Click on the tree entry which appears afterwards and use the popup
menu (invoked by the right mouse click) to show the tree.
Check out what happens when you right click on the tree viewer panel
(on inner nodes, on leaf nodes and on the background)
Offline Swissprot database access
In order to use the option to view local copies of Swissprot entries you
will have to run a small Java application included in the jar/tar file.
This application reads a Swissprot database file and splits it into pieces.
Each piece is stored in a separate file beneath the base directory outdir.
A single piece is defined by a range within a file starting after a line with a "//" prefix and ending
before the next line with a "//" prefix.
The exact path to each file is derived from the accession
number of the corresponding Swissprot entry.
In order to work properly there must be a line for each piece / Swissprot entry
starting with the letters AC, followed by three
spaces and a string with at least 6 characters. Assuming these 6 characters are
123456, the resulting filename will be "outdir/1/23/123456.spr"
If you run the application using the following command java org.eml.trees.SplitFile , e.g. java org.eml.trees.SplitFile xxx /usr/data/sprot/
where xxx is the whole Swissprot file,
you will get a directory structure like
where a protein entry is e.g. stored in a file like /usr/data/sprot/O/01/O01258.spr
This is the kind of file being looked for when accessing a text file
from a leaf node shown in TreeWiz. You will have to specify the Swissprot base dir in the Properties file.