When I use Louvain or Infomap for Community Detection in Cyto I get a Community Graph instead of a "regular" nodes' graph with nodes colored by community association.
It's difficult to analyze the community graph without knowing which nodes are in which communities.
The grid that shows the nodes' ID for each community doesn't contain info about the nodes.
The community graph can be useful only when I understand what's the community about (like after using WordCloud to identify the community traits).
Thank you for your time and for the plugin!

I use Win10, Cyto ver:3.8.2

It was pointed out that the order of custom parameters set for a given algorithm in communitydetectionalgorithms.json configuration file is not preserved in the GUI. Ideally that order should be kept cause the developer may not certain arguments appearing before others.

For networks that are NDEx normalized gene symbols for complexes are stored in the attribute members This causes a problem because the CD_MemberList only contains node names so these genes are lost in functional enrichment.

Example network:
http://ndexbio.org/#/network/077e15fd-1e5f-11e8-b939-0ac135e8bacf

gProfiler already reports pValues that are adjusted for multiple hypothesis testing (within each node). It might be nice to be able to set the pValue based on multiple hypothesis testing for the network, i.e. Benjamini-Hochberg, Storey, or Bonferroni based on the number of nodes. Not essential, as we could always calculate a conservative Bonferroni based simply on the number of nodes ourselves, but this is just something to consider.

It would be nice if one could dictate the maximum number of edges/nodes a given community detection algorithm could support.

To reproduce using version 1.10.0

#. Open a network and invoke Apps -> Community Detection -> Run Community Detection
#. Pick "Louvain" and change Cluster Resolution to another value say 11
#. Click either Run or Cancel button.

#. Invoke Apps -> Community Detection -> Run Community Detection
#. Pick "Louvain" and note that Cluster Resolution has changed back to its original value

Using version 1.10.0 Ran into an issue where gprofiler was being run on a large number of nodes and one of the calls failed with this error message:

The progress dialog then disappeared, but the network kept flickering and other calls to gprofiler continued to be process with no way for user to cancel.

The App should either stop all further processing and close progress bar or somehow denote the failure for given node and continue.

Hi all, I was exploring your tool in cytoscape and am impressed with its ease of use. One question that I have is how many iterations of OSLOM are run by default? Looking at the OSLOM documentation, the default is 10 (-r flag), but I don't see that mentioned anywhere in your documentation. Thanks for clarifying!

If one creates a hierarchy via this tool and then renames the CD_MemberList_Size and then performs term mapping they will get a null pointer exception dialog like this one:

The code assumes this column exists. At the minimum a more friendly error message telling the user an expected column is missing should be shown instead of the exception.

Raw log:

java.util.concurrent.ExecutionException: java.lang.NullPointerException: Cannot invoke "java.lang.Integer.intValue()" because the return value of "org.cytoscape.model.CyRow.get(String, java.lang.Class)" is null
	at java.util.concurrent.FutureTask.report(FutureTask.java:122) ~[?:?]
	at java.util.concurrent.FutureTask.get(FutureTask.java:191) ~[?:?]
	at org.cytoscape.app.communitydetection.termmap.TermMappingTask.run(TermMappingTask.java:102) ~[?:?]
	at org.cytoscape.work.internal.task.JDialogTaskManager$TaskRunnable.innerRun(JDialogTaskManager.java:321) ~[!/:?]
	at org.cytoscape.work.internal.task.JDialogTaskManager$TaskRunnable.run(JDialogTaskManager.java:352) [!/:?]
	at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:539) [?:?]
	at java.util.concurrent.FutureTask.run(FutureTask.java:264) [?:?]
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136) [?:?]
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635) [?:?]
	at java.lang.Thread.run(Thread.java:833) [?:?]
Caused by: java.lang.NullPointerException: Cannot invoke "java.lang.Integer.intValue()" because the return value of "org.cytoscape.model.CyRow.get(String, java.lang.Class)" is null
	at org.cytoscape.app.communitydetection.termmap.TermMappingCallable.call(TermMappingCallable.java:149) ~[?:?]
	at org.cytoscape.app.communitydetection.termmap.TermMappingCallable.call(TermMappingCallable.java:20) ~[?:?]
	... 4 more
java.util.concurrent.ExecutionException: java.lang.NullPointerException: Cannot invoke "java.lang.Integer.intValue()" because the return value of "org.cytoscape.model.CyRow.get(String, java.lang.Class)" is null
	at java.base/java.util.concurrent.FutureTask.report(FutureTask.java:122)
	at java.base/java.util.concurrent.FutureTask.get(FutureTask.java:191)
	at org.cytoscape.app.communitydetection.termmap.TermMappingTask.run(TermMappingTask.java:102)
	at org.cytoscape.work.internal.task.JDialogTaskManager$TaskRunnable.innerRun(JDialogTaskManager.java:321)
	at org.cytoscape.work.internal.task.JDialogTaskManager$TaskRunnable.run(JDialogTaskManager.java:352)
	at java.base/java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:539)
	at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:264)
	at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136)
	at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635)
	at java.base/java.lang.Thread.run(Thread.java:833)
Caused by: java.lang.NullPointerException: Cannot invoke "java.lang.Integer.intValue()" because the return value of "org.cytoscape.model.CyRow.get(String, java.lang.Class)" is null
	at org.cytoscape.app.communitydetection.termmap.TermMappingCallable.call(TermMappingCallable.java:149)
	at org.cytoscape.app.communitydetection.termmap.TermMappingCallable.call(TermMappingCallable.java:20)
	... 4 more

If nodes in the parent network have spaces their names, those nodes will not be displayed when a user clicks on a hierarchy network node and runs View interactions for Selected Node

This can be seen with single node hierarchy in this cytoscape 3.8 session (gzipped cause github does not like .cys extension):
nodeswithspaces.cys.gz

Test network that was created

Single node hierarchy created with OSLOM:

When a user updates the REST Server (app.baseurl) in the Settings menu found by clicking on Apps => Community Detection => Settings the App attempts to query that new server to get a list of algorithms. This can take several seconds and during this time Cytoscape appears unresponsive.

It would be nice to provide a progress dialog denoting this operation is taking place (with cancel option)

For example in Louvain:

The checkboxes Generate overlapping communities and Generate deep hierarchy cannot both be checked, it can be none or one of them.

When running Functional Enrichment the symbols in CD_MemberList are sent to the user selected enrichment algorithm. This list is derived from the name column in the parent network which may not contain valid gene symbols. It would be nice to let the caller dictate an alternate column.

If you create a node only network you can run community detection algorithms, but this should be disallowed since the data sent is an edge list.

These should be discussed and broken into tickets:

• Column selection for weight can only take positive values where all values are set

• Better warnings on column weight

• People need help interpreting result they got, look at gene lookup in hiview (smart pubmed searches)

• Support multiple organisms (term mapping)

• Enable export to hiview