"Coral: an integrated suite of visualizations for comparing clusterings"


To run: java -Xmx1G -jar coral.jar


To use:

  1. Load your clusterings through File->Load clusterings...
     Your-clustering-name.txt files should have the following format:

       module1 item-1 item-2 item-3
       module2 item-3 item-4 item-5 item-324
     
  2. After your data has loaded, you can reorder the co-cluster matrix
     through Matrix->Reorder... You can specify how many iterations of
     greedy LAP and of the Hungarian optimal LAP you want to run.

  3. If your modules were generated form a network, you can upload the
     network edg file to show cohesion. File format should be:

       item-1,item-2
       item-1,item-3
       ...

  4. You can add/remove column from item pairs and module tables by
     going to View->Items table and View->Module table repesctively     


Data:

  arabidopsis
    9 clusterings described in the paper (see above). Each clustering
    file has a name of a clustering algorithm that was used to 
    generate the clustering. A.thal.edg file contains the PPI network
    reported in:
      Vidal et al. "Evidence for Network Evolution in an Arabidopsis
      Interactome Map." Science, 333(6042):601-607. 2011.

  yeast
    4 clusterings of yeast synthetic lethal iteractions network
    generated for:
      Kelley, Kingsford. "Extracting Between-Pathway Models from 
      E-MAP Interactions Using Expected Graph Compression" Journal of
      Computational Biology. 18(3):379-390. 2011.

    bandyo - Bandyopadhyay, S., Kelley, R., Krogan, N.J., et al. Functional maps
      of protein complexes from quantitative genetic interaction data. PLoS 
      Comput. Biol. 2008. 4(4): e1000065.
    us - Ulitsky, I., Shlomi, T., Kupiec, M., et al. From E-MAPs to module
      maps: dissecting quantitative genetic interactions using physical 
      interactions. Mol. Syst. Biol. 2008. 4: 209.
    arthur - Brady, A., Maxwell, K., Daniels, N., et al. 2009. Fault tolerance 
      in protein interaction networks: stable bipartite subgraphs and redundant 
      pathways. PLoS ONE 4, e5364.
    pgs    - expected (probabilistic) graph summarization (Kelley)

  karate_club
    10 near-optimal clusterings of the karate club dataset. Clusterings
    were generated by Clauset's modularity method for Navlakha, Kingsford
    study:
      Navlakha, Kingsford. "Exploring Biological Network Dynamics with
      Ensembles of Graph Partitions". PSB, 2010.


Website: http://www.cbcb.umd.edu/kingsford-group/coral/

If you have any comments or questions, email dfilippo@cs.umd.edu
