Repeated Random Walk
Implementation Download


An implementation for the clustering algorithm, Repeated Random Walks (RRW) can be downloaded here. The program is written in Java, and takes a tab-deliminated file representing a weighted graph as input. The graph file should follow the format:

NodeID<tab>NodeID<tab>LinkWeight<newline>
NodeID<tab>NodeID<tab>LinkWeight<newline>
etc.

To run the program, the usage follows:

java RRW -i inputGraph [-o outputFile] [-r restartProb] [-max maxClusterSize] [-min minClusterSize] [-overlap overlapThreshold] [-lambda earlyCutoff]

Default values for the various options are as follows:
outputFile = RRW_output.txt
restartProb = 0.7
maxClusterSize = 11
minClusterSize = 5
overlapThreshold = 0.2
earlyCutoff = 0.6

Download: RRW.tar.gz

The yeast functional network used in the paper can be found from the supporting materials section of (Keimer et al. 2007) available here.

The three noisy networks used in the paper are available for download here:
FN40
FP40
Rewire40

The resulting clusters from running the RRW algorithm on the WI-PHI network is available for download here:
RRW Clustering for the WI-PHI network

The MIPS annotations for yeast can be downloaded from the FTP section of the MIPS Comprehensive Yeast Genome Database found here.

The GO annotations for yeast can be downloaded from the Annotation Downloads section of the Gene Ontology website found here.