- If not installed yet, java 6 or higher should be installed
- Download the Diffany library jar file and add it to your classpath
- Download the Diffany cli jar file
- The cli jar is intented for generating text-readable differential networks
- Within one root input directory, the reference network and condition-dependent network(s) need to be defined in separate subdirectories each, containing these .txt files:
- network.txt: contains the ID, name and type of the network, as well as its pre-defined node attributes
- nodes.txt: contains one node per line, specifying the node ID and symbol, as well as the values for the node attributes
- edges.txt: contains one edge per line, in a tab-delimited format: source_ID \t target_ID \t interaction_type \t symmetrical/directed \t weight \t negated/affirmative
- conditions.txt: only needed for the ConditionNetwork. Specifies one condition per line, consisting of its description and an (optional) list of tab-delimited ontology terms
- In a similar format as described above, the differential and the overlapping networks will each be written to a separate output directory with the same .txt files.
-
Basic functionality: java -jar Diffany_CL_1.0.0.jar -i <inputdir> -o <outputdir>
-
Run the jar with -v or --version (only!) to obtain the version number of the Diffany tool
-
Run the jar with -h or --help (only!) to obtain the help file with more detailed instructions:
-
usage: java -jar Diffany_CL_1.0.0.jar [-c <arg>] [-cons <arg>] [-diff <arg>] [-h <arg>] -i <dir> [-ID <arg>] [-l] [-m <arg>] -o <dir> [-oper <arg>]
- -c,--confidence <arg> the minimum confidence threshold for output edges, as an integer or double (default=0.0)
- -cons,--consensus <arg> whether or not to calculate consensus networks: yes or no (default=yes)
- -diff,--differential <arg> whether or not to calculate differential networks: yes or no (default=yes)
- -h,--skipHeader <arg> whether or not to skip the first line (header) in the network .txt files (default=yes)
- -i,--inputDir <dir> the input directory containing the reference and condition-specific networks
- -ID,--outputID <arg> the first ID that will be used for the generated networks
- -l,--log display a progress/log file during the run
- -m,--mode <arg> the mode of comparison: pairwise or all (default=all)
- -o,--outputDir <dir> the output directory which will contain the generated differential/consensus networks
- -oper,--operator <arg> the operator used to create consensus edges: min or max (default=min)