SLiMScape is developed and maintained by Kevin O'Brien, a PhD student in the Shields lab in University College Dublin, Ireland.

Download:

Plugin:

Latest Version
SLiMSuite.zip

Example Data:

Session.cys
slims.csv

Jump to the Getting Started for instruction on how to load sample data or jump to the tutorial for an example analysis.

Table of Contents

Getting Started:

To install SLiMScape you just need to place SLiMScape.jar in your Cytoscape plugins directory.

  1. Load some example data and launch The plugin
    1. launch Cytoscape

    2. Open session.cys (File->open).

    3. Click "plugins" menu and then "launch SLiMScape"
      Launch screen
  2. A popup will tell you that this is the first time you have run SLiMScape. Press OK and you will be brought to a config page where you need to enter the path to a writeable directory. This directory will be used to store results. The config screen can be seen below. Most of the options can be ignored unless you want to run SLiMFinder and SLiMSearch locally.
    Options screen

Then click save. This will save these settings in slimscape_properties.txt in the Cytoscape folder.


  • Launch screen
  • Now switch back to the "Import Network" tab. You must select the node label attribute and the uniprot attribbute which must contain a label which is suitable for naming each node and a uniprot artrtribute, respectively.
  • Click import and wait for the data to load.

There is a tutorial on how to perform a basic analysis here

Using the Plugin:


Introduction:

The root SLiMScape control panel can contain up to five tabs depending on whether you have imported custom attributes and Domains. There are two types of panel.


  1. Analysis:

    Analysis panels are initially empty and are populated each time SLiMFinder or SLiMSearch are run on a set of nodes. The result of these are called called "Runs" and they are stored in the "Runs table". Each non empty result in the results table links to a set of SLiMFinder or SLiMSearch results and can be displayed in the "Results Table" by clicking on a specific run.

    1. SLiMFinder Panel
      • Provides an interface to run SLiMFinder and discover new motifs in the protein interaction network. Proteins are found by searching for statistically overrepresented sequences withoin a set of proteins.
    2. SLiMSearch Panel
      • Provides an interface to run SLiMSearch and identify new instances of a predefined motif in the protein interaction network. Regular expressions are used to specify the motif.
  2. Support:

    Support panels provide supporting evidence for the existence or function of a SLiM. These consist of domains and user defined custom attributes.

    1. Domain Panel
      • Provides a list of domains within the network. Each entry links to a single node and therefore there are many instances of the same domain type in this list.
    2. Aggregate Domain Panel
      • This contains the same data as "Domain Panel" but duplicate values are merged and therefore each domain matches to one or more nodes. Some analysis functionality is also available, as SLiMFinder sets can be defined from protein with a specific domain or which interact with a specific domain.
    3. Custom Attributes Panel
      • Contains a list of user defined attributes which can be visualized in the network. Each attribute matches to one node.
Root SLiMScape Control Panel



Overview:

Each tab presents SLiMs, domains or custom attributes within a table and each entry within these tables can be selected. Selected entries are highlighted within the network, as shown in example 1.


Example 1: Four instances of a SLiM are highlighted in red diamonds and two instances of the RQC domain are highlighted as blue squares.

The colour codes are shown below:

Icon Description
Default node icon
Run colour (Can have other shapes)
SLiM icon
Domain icon
SLiM and Domain icon

Input:

The minimum input for SLiMScape is an interaction network which contains a Uniprot attribute for each node. When launching SLiMScape you must indicate the column which contains the Uniprot attribute. This will be used to retrieve protein sequences from Uniprot and these sequences will be used for the SLiM analyses and retrieval of domains from the SMART database.

Nodes: 
ID1,ID2
O14924,P05106
P05106,O35431
P05106,P07384
P05106,P08123
P05106,B9A6L2
P05106,P21980
P05106,Q13485
P05106,P54939

Additionally, you can import a network with non Uniprot IDs and import the Uniprot IDs as attributes as shown below:

Nodes: 
ID1,ID2
RGS12,ITGB3
ITGB3,Apba2
ITGB3,CAPN1
ITGB3,COL1A2
ITGB3,RABGAP1
ITGB3,TGM2
ITGB3,SMAD4
ITGB3,TLN1

Attributes: 
ID,uniprot
ITGB3,P05106
RGS12,O14924
Apba2,O35431
CAPN1,P07384
COL1A2,P08123
RABGAP1,B9A6L2
TGM2,P21980
SMAD4,Q13485
TLN1,P54939



SLiMFinder:

The SLiMFinder panel acts as an interface for the SLiMFinder executable and allows automated or manual analysis of the network. To run SLiMFinder you must first select a "Run Configuration" from the drop down menu. A "Run Configuration" simply defines a method for grouping nodes which will be analyzed using SLiMFinder. Details of each run configuration can be seen below.

slimfinder interface
Each of these configurations defines on or more sets of proteins. Each of these sets is then separately analyzed using SLiMFinder.
  1. Selected nodes
    • This will run SLiMFinder on only the selected nodes. This can be useful if you have a very specific set of proteins to analyze.
  2. Batch Interactions (spokes + hub)
    • This will create several sets of nodes and SLiMFinder will be run once for each set. Each set is centred on a single protein and its interactors (spokes). This is useful to discover SLiMs which are shared between and set of proteins and may mediate interactions.
  3. Batch Interactions (spokes)
    • This is the same as the "Batch Interactions (spokes + hub)" configuration but the hub protein is excluded.
  4. Selected Node Interactors (hub)
    • Sets are defined as the interactors of the currently selected nodes. One set is created per selected node.
  5. Batch Domains
    • This will create several sets of nodes and SLiMFinder will be run once for each set. Each set is a group of proteins that contain a particular domain.
  6. Selected Domain Interactors
    • This will create several sets of nodes and SLiMFinder will be run once for each set. Each set is a group of proteins that interact with a particular domain.
  7. Selected Node Interactors (spokes + hub)
    • Sets are defined as the interactors of the currently selected nodes. One set is created per selected node.
  8. Merge Selected Node Interactors (hub)
    • A single set is defined consisting of the shared interactors of the currently selected nodes.
  9. Merge Selected Node Interactors (spokes + hub)
    • A single set is defined consisting of the shared interactors of the currently selected nodes.
  10. Attribute
    • Multiple sets are defined based on the value of a node attribute. This may be useful for finding targeting proteins. For example, using a sub cellular location attribute may find targeting motifs for specific regions of the cell. Other approaches are also possible depending on the chosen attribute.

After selecting a run configuration, click "Run SLiMFinder". This will create one or more sets of protein sequences and execute SLiMFinder.


SLiMFinder Options

SLiMFinder is a complicated software package with many command line options. SLiMScape presents a simplified interface for these options and for most users the default values will be sufficient.

slimfinder interface
Masking
  • Disorder Masking is used to mask residues which have an IUPred disorder score of less than 0.3
  • Conservation Masking is used to mask residues which have a relative local conservation score of less than X. You must provide an orthDB field in the startup options panel in order to use this function.
  • Feature Masking is used to mask residues which occur in features such as domains or transmembrane regions.
SLiMChance
  • Probability Cutoff is the cutoff for returned motifs.
Miscellaneous Options
  • Walltime is the maximum runtime of a single run.
  • Custom parameters is used to add other command line arguments which can be found in the SLiMFinder documentation.

SLiMFinder Results

SLiMFinder Interface

SLimFinder results are displayed in the "Results View" in the lower left panel of the "Run SLiMFinder" tab. The "Runs View" in the middle panel contains one entry per SLiMFinder run. When a row is clicked either the log or the results are displayed in the results tab. Clicking the checkbox in the "Runs Table" will highlight all nodes which are included in the run in grey. Selecting a SLiM in the results view will highlight all nodes which contain that SLiM in red.



SLiMSearch:

The SLiMFinder panel acts as an interface for the SLiMSearch executable. Unlike SLiMFinder there is only one run configuration where only selected nodes a are searched. To run SLiMSearch you must provide one or more regular expressions in the motifs box and select one or more nodes. This will search the protein sequences of the selected nodes for occurrences of the specified regular expressions. This is useful for locating new instances of a motif found using SLiMFinder.

slimfinder interface

SLiMSearch Options

slimfinder interface
Masking
  • Disorder Masking is used to mask residues which have an IUPred disorder score of less than 0.3
  • Conservation Masking is used to mask residues which have a relative local conservation score of less than X. You must provide an orthDB field in the startup options panel in order to use this function.
SLiMChance
  • Probability Cutoff is the cutoff for returned motifs.
Miscellaneous Options
  • Walltime is the maximum runtime of a single run.
  • Custom parameters is used to add other command line arguments which can be found in the SLiMSearch documentation.

SLiMSearch Results

SLiMSearch Interface

SLimFinder results are displayed in the "Results View" in the lower left panel of the "Run SLiMFinder" tab. The "Runs View" in the middle panel contains one entry per SLiMFinder run. When a row is clicked either the log or the results are displayed in the results tab. Clicking the checkbox in the "Runs" will highlight all nodes which are included in the run in green. Selecting a SLiM in the results view will highlight all nodes which contain that SLiM in red.


Deleting Results:

Invalid, empty or cancelled results can be removed by right clicking on an entry and pressing the "Purge Results" button. This will remove them from the list and deactivate any active entries which they contain. Non empty results can be deleted in the same way if they are first cancelled. To do this right click on the relevent row and click "Cancel/Mark for deletion".


Domains:

Ligand binding SLiMs often interact with their binding partner via a structured domain. Common examples include SH2, SH3 and PDZ domains. This makes them an important indicator when attempting to determine the function or existence of a SLiM. SLiMScape retrieves proteins domains for each node using the SMART[4]⁠ webservice.

Domain Results:

slimfinder interface

Domain Analysis:

Domains can be used to define input sets for SLiMFinder by right clicking on any domain and selecting either "SLiMFinder: Single domain" or "SLiMFinder: Single domain interactors" as shown in the image above.

  • SLiMFinder: Single domain
    • Run SLiMFinder on all proteins which contain the specified domain.
  • SLiMFinder: Single domain interactors
    • Run SLiMFinder on all interactors of proteins which contain the specified domain. This strategy has previously been used to define a set of sequences which bind to specific domains. (see slimdb)


Custom Attributes:

Custom attributes allow user defined slims or other features to be visualized. The contents of these attributes is completely optional.

To import custom attributes you must select the "Import Custom Attributes" checkbox when launching the plugin. Custom Attributes. These must be supplied as a separate file with the following format, however extra column are permitted:

Node_id_title,attribute_name_title
id1,name
id2,name

Running SLiMFinder and SLiMSearch locally:

You will also need to download and install the following software:

Cytoscape 2.8.3
Python 2.7.3
ClustalW2 - version 2.0.12
Legacy BLAST - version 2.2.26
IUPRED - version 1.0
Muscle - version 3.8.31

To install SLiMScape you just need to place SLiMScape.jar in your Cytoscape plugins directory.

  1. Unzip slimsuite.zip
  2. Load some example data and launch The plugin
    1. launch Cytoscape

    2. Open session.cys (File->open).

    3. Click "plugins" menu and then "launch SLiMScape"
      Launch screen
  3. A popup will tell you that this is the first time you have run SLiMScape. Press OK and you will be brought to a config page where you will be asked to provide the following information:
    Options screen

Root Options:

Option Description Type Optional
Python executable This should point to the python 2.7 executable. Note, if you select "Detetc Automatically" python 2.7 exec must be in the PATH environment variable File ✗
slimsuitehome This should point to the slimsuite directory. Directory ✗
clustalwpath This should point to the clustalW2 executable. File ✔
orthdbpath This should point to an orthologue database. File ✔
outputdirectory This is where results are stored. It should be set to any writable directory. Directory ✗
musclepath This should point to the muscle executable. File ✔
blastpath SThis should point to the blast/bin directory. Legacy blast must be used. Blast+ will not function. Directory ✗

Sample OrthDb's are available from data@bioware


Then click save. This will save these settings in slimscape_properties.txt in the Cytoscape folder. An example of the contents of this file is as follows:

#Wed Oct 10 11:37:12 IST 2012
slimsuitehome=/home/kevin/Desktop/slim_domain_vis/slimsuite
clustalwpath=/home/kevin/bio_programs/Tools/clustalw2/clustalw2
orthdbpath=/home/kevin/Desktop/slim_domain_vis/proteomes/ens_loci_proteomes.101018.fas
outputdirectory=/home/kevin/Desktop/slim_domain_vis/slimsuite/runs/
pythonhome=
musclepath=/home/kevin/bio_programs/Tools/muscle/muscle
blastpath=/home/kevin/bio_programs/Tools/blast/bin
uipredpath=/home/kevin/bio_programs/Tools/iupred/iupred

Troubleshooting:

  • Repeated errors when running SLiMFinder locally on Windows.
    • This could be caused by a bug when calling blast. If the path to the blast executables contains any spaces, the command may not execute correctly. To fix this, simply move the blast directory to a path with no spaces (the root of the C drive for example) and update the blast path within the plugin.
  • Conservation Masking has been disabled on the SLiMFinder and SLimSearch options screens when running SLiMsuite locally.
    • This happens when the orthdb file or clustalw executable do not exist. To fix this please provide valid paths to these files in the root options panel.
  • Disorder Masking has been disabled on the SLiMFinder and SLimSearch options screens when running SLiMsuite locally.
    • This happens when the IUPred executable does not exist. To fix this please provide valid paths to this file in the root options panel.
  • While running SLiMFinder or SLiMSearch locally there are repeated popups indicating that blast has crashed.
    • This could be caused by running SLiMScape with insifficient RAM.

Contact:

Please email if you have any queries.


Links:

Legacy BLAST - version 2.2.26
ClustalW2 - version 2.0.12
IUPRED - version 1.0
Muscle - version 3.8.31
SLiMFinder
SLiMSearch
Bioware Server