CDAPS¶

Community Detection APplication and Service (CDAPS)

CDAPS performs multiscale community detection and functional enrichment for network analysis through a service-oriented architecture. These features are provided by integrating popular community detection algorithms and enrichment tools. All the algorithms and tools run remotely on a dedicated server.

Currently supported features:

Community detection algorithms: Louvain, Infomap, OSLOM, CliXO, HiDeF
Functional enrichment tools: g:Profiler, Enrichr, iQuery

Feature Requests and Reporting Bugs¶

The CDAPS GitHub issue tracker can be used to report a bug or request a feature.

To Report a bug:

Go to https://github.com/cytoscape/cy-community-detection/issues
Click on New Issue
Write a short description of the issue. It is very helpful to provide a series of steps that can be taken to reproduce the issue.
If possible attach a session file (.cys) or example input files.
Enter App version, Cytoscape version and operating system.
Click on Submit new issue

Cite CDAPS¶

Singhal A, Cao S, Churas C, Pratt D, Fortunato S, Zheng F, et al. (2020) Multiscale community detection in Cytoscape. PLoS Comput Biol 16(10): e1008239. https://doi.org/10.1371/journal.pcbi.1008239

Installation¶

Cytoscape minimum version 3.7 is required.

Install Cytoscape

Download and install the latest version of Cytoscape at
https://cytoscape.org/download.html

Install CyCommunityDetection

Open Cytoscape

In the main menu select Apps -> App Manager

In the App Manager select CyCommunityDetection in the list of All Apps and click the Install button.

CyCommunityDetection can also be installed from the Cytoscape App Store at
https://apps.cytoscape.org/apps/cycommunitydetection

What’s New¶

Version 1.12.0¶

Added new menu option Apps -> Community Detection -> Tally Attributes in Hierarchy that annotates the hierarchy network by adding columns using values tallied from a set of attributes/columns in the parent network.
Order of algorithms displayed in Run Community Detection and Run Functional Enrichment dialogs are now consistent with what is returned from CDAPS REST server. UD-1573

Version 1.11.0¶

Added support for new edge list format COMMUNITYDETECTRESULTV2 allowing Community Detection algorithms to add additional custom node columns. UD-1091

Bug fixes

Fixed bug where View Interactions for Selected Node no longer worked if a session was reloaded. UD-1087
Fixed bug where any values set in the community detection or functional enrichment dialogs were lost when when reopening those dialogs. Issue #3

Version 1.10.0¶

Added new menu option Apps -> Community Detection -> Settings that lets caller easily change CDAPS REST server. UD-1066
Added message letting user know using weight column in Community Detection dialog is an advanced parameter. UD-988
In Community Detection dialog replaced About button with info icon next to algorithm selection dropdown. UD-987

Bug fixes

Fixed bug where changes to properties under Edit -> Properties -> CyCommunityDetection were not being loaded. UD-986

Version 1.0¶

First release

Quick Tutorial¶

Open a network¶

To run Community Detection, a network must be loaded in Cytoscape.

From within Cytoscape click on Affinity Purification network on the starter panel:

If not displayed, the Starter Panel can be displayed by invoking the menu option View -> Show Starter Panel.

Run Community Detection¶

With the network loaded click on Apps -> Community Detection -> Run Community Detection menu option.

The above step will display a dialog seen below.

Select Louvain from algorithm dropdown and click Run button.

A new network/hierarchy should be generated as seen here (The current default layout will be used):

Note

Each node in the network/hierarchy above represents a cluster with the members of that cluster set in the CD_MemberList node column

Perform Term Mapping¶

Using network/hierarchy generated above select a few nodes in the network view and then right click on a selected node to display the submenu and select Apps -> Community Detection -> Run Functional Enrichment as seen here:

The above step will display a dialog seen below:

Select gProfiler from algorithm dropdown and click Run button.

gProfiler will be run and nodes will be named and colored according to overlap as seen here:

View Interactions¶

Using network/hierarchy generated above select a single node in the network view and then right click on the selected node to display the submenu and select Apps -> Community Detection -> View Interactions for Selected Node

Invoking this menu option will show all the nodes pertaining to this cluster in the parent network like as seen here:

Tally Attributes on Hierarchy¶

With a CDAPS hierarchy network (as generated via Run Community Detection step above) click on Apps -> Community Detection -> Tally Attributes on Hierarchy menu option.

The above step will display a dialog seen below.

Select one or more columns from parent network to tally on hierarchy network and click OK button. (only columns of type Integer, Boolean, or Double will be listed)

New corresponding columns will be added to the hierarchy network node table as seen below.

These new columns match the names of the columns selected in the dialog, but with the namespace CommunityDetectionTally prefixed.

In addition, a CommunityDetectionTally::Unmatched column will also be added.

Note

These columns can be used to create charts on each node as described here

Example hierarchy network styled with node charts:

For more information click: Tally Attributes on Hierarchy

Tally Attributes on Hierarchy¶

The menu option Apps -> Community Detection -> Tally Attributes in Hierarchy provides a way to annotate the hierarchy network created when running a Community Detection algorithm from this App.

Example of Apps -> Community Detection -> Tally Attributes in Hierarchy styled with node charts

More specifically, Apps -> Community Detection -> Tally Attributes in Hierarchy menu option provides a way to count the number of members in each hierarchy network node that have a true/positive value for a specified set of attributes/columns in the parent network.

Note

Members of of a hierarchy node cluster are set in the node CD_MemberList column as a space delimited string of parent node names

These counts are stored as new columns/attributes on the hierarchy network with the same name as seen in the parent network, but prefixed with CommunityDetectionTally namespace.

In addition, any members in the hierarchy cluster that do NOT match any of the specified set of attributes/columns are counted in the Unmatched column/attribute.

Warning

For attribute(s)/column(s) of type Double, the value is rounded to nearest integer before checking to see if the value is positive

To Run¶

Select a hierarchy network created by CDAPS in Cytoscape and click on Apps -> Community Detection -> Tally Attributes on Hierarchy menu option.

Note

When run, any existing columns in CommunityDetectionTally namespace will be removed from the hierarchy network.

Columns¶

This page describes the columns created by CDAPS in the network and node tables.

Network Columns¶

Columns created in network table when invoking Run Community Detection

name - String in format:

Example:

__CD_OriginalNetwork - SUID of parent network. This value can change upon session save/reload. If CDAPS is unable to find the parent network a dialog will be displayed to the user letting them select the correct parent network.

Note

Setting the value of __CD_OriginalNetwork network attribute to a negative value will display the Parent network dialog chooser the next time View Interactions is invoked

description - Contains string in this format:

Example:

prov:wasDerivedFrom - Name of parent network

prov:wasGeneratedBy - Denotes version of CDAPS and algorithm used in this format:

Example:

Node Columns¶

In the node table columns created by CDAPS are prefixed with CD_

Columns created in node table when invoking Run Community Detection

CD_MemberList - String of space delimited node names representing members of this cluster

CD_MemberList_Size - Size of CD_MemberList

CD_MemberList_LogSize - Log of CD_MemberList_Size

CD_CommunityName - Name of community set by invocation of Run Functional Enrichment

CD_AnnotatedMembers - String of space delimited node names used to set value in CD_CommunityName

CD_AnnotatedMembers_Size - Size of CD_AnnotatedMembers

CD_AnnotatedMembers_Overlap - CD_AnnotatedMembers_Size divided by CD_MemberList_Size

CD_AnnotatedMembers_Pvalue - Pvalue obtained from term mapping algorithm invoked by Run Functional Enrichment

CD_Labeled - Boolean denoting if CD_CommunityName was set to a value other then blank or (none)

Columns created in node table when invoking Run Functional Enrichment

CD_AnnotatedAlgorithm - Algorithm used to set value in CD_CommunityName in format:

Example:

CD_NonAnnotatedMembers - String of space delimited node names NOT used by algorithm to set mapped term in CD_CommunityName

CD_AnnotatedMembers_SourceDB Source database used by algorithm to set mapped term in CD_CommunityName

CD_AnnotatedMembers_SourceTerm Id of mapped term set in CD_CommunityName

Columns created in node table when invoking Tally Attributes on Hierarchy

CommunityDetectionTally::<TALLY SELECTED COLUMNS> - Each column selected in the tally will have an entry and the value is the count of members that did have a positive or true value in the column on the parent network

CommunityDetectionTally::Unmatched - Count of members in hierarchy cluster node that did NOT have a positive or true value in any of the tally columns

Note

CommunityDetectionTally is the namespace prefixed onto columns created when invoking Tally Attributes on Hierarchy

For more information click: Tally Attributes on Hierarchy

Settings¶

CDAPS allows user configuration via two methods.

The first method¶

User configurable options can be changed via a dialog built into CDAPS. This can be found by clicking on the Cytoscape menu as follows:

Apps -> Community Detection -> Settings

Doing the above will display a dialog as seen here:

Currently the only setting that can be changed is to set an alternate CDAPS REST Server which is the server that runs the community detection and functional enrichment (term mapping) algorithms.

Note

To reset back to the default server just remove all text from this field and click Update

The second method¶

Warning

This is for advanced users only and may cause CDAPS to fail and/or to use excessive CPU and network bandwidth

Edit -> Preferences -> Properties

In the dialog that appears, CDAPS configuration can be found by selecting CyCommunityDetection from dropdown menu.

Frequently Asked Questions¶

How to set alternate parent network¶

To set an alternate parent network for a hierarchy, change the value of the network attribute __CD_OriginalNetwork to a negative number such as -1 as seen here:
Load the alternate parent network into Cytoscape. Then select a node in the hierarchy and invoke View Interactions to, if needed, display the Parent network dialog as seen below:

Note

Parent network dialog will only be displayed if there is more then one potential parent network

How to make a network look like a hierarchy network¶

To make hierarchy network from scratch, do the following:

Create/load a network and add nodes and edges to represent the hierarchy
Add __CD_OriginalNetwork network attribute column with type set to Long Integer and the value set to -1. This column tells this tool the network is a hierarchy.
Add CD_MemberList node column to each node in network with type String and with the value set to a list of space delimited node names that correspond to name column values from the parent network. This column is needed when doing View Interactions
Add CD_MemberList_Size node to each node in hierarchy with type Integer and with the value set to the number of values in CD_MemberList node column. This column is needed when doing Term Mapping
If not already done, load the parent network into Cytoscape. In screenshot above, HIV-human PPI network has been loaded

Note

To get the same look and feel, copy the style from a hierarchy already run by this tool onto the hierarchy made from scratch