Menu

Use Case 1 HOW TO FIND NEGATIVE REGULATORS OF A SPECIFIC GENE

Biological question

Which Transcription Factors (TFs) and microRNAs (miRs) inhibit PTEN and can be investigated as a cancer therapy target?

UseCase1

The protein PTEN is encoded by a tumor suppressor gene that is mutated in a large number of cancers, for example, head and neck (PMID: 11801303) and glioma cancer (PMID: 12085208). Regulation of the protein level can be done at the gene transcription level, by Transcription Factors, or post-transcriptionally, for instance by miRNA inhibition. To find the regulators at both levels, the Cytoscape software can be used as a platform to display the information of different resources and databases on this matter. General instructions about how to use this software and the software apps are available at //cytoscape.org/

SOLUTION

  1. Use Cytoscape to identify those miRNAs that regulate PTEN at the mRNA level, by importing miR:mRNA interactions from PSICQUIC. This resource retrieves information related to molecular interactions from different databases or active services; including IntAct and GO resources.
  2. Use Cytoscape to identify those TFs that regulate PTEN, by importing TF:TG interactions from PSICQUIC. 

The query can be done initially in the PSICQUIC View web page, downloaded and then imported into Cytoscape as a network, or the results can be directly seen as interaction networks using the PSICQUIC core app in Cytoscape.

By going into the PSICQUIC web page, an initial investigation can be performed to check the data associated with PTEN available in Ensembl and UniProt, via the respective IDs.

  • To find all interactions with PTEN and genes, paste the following query into IntAct or PSICQUIC (adapt to your protein of interest): (ENSG00000171862) AND ptypeA:protein AND ptypeB:gene

From the initial query, the interactions from the different services can be downloaded and imported as interaction networks in Cytoscape.

arrowblue

By searching directly in the Cytoscape plugin, the Interactor ID needs to be included and the data sets from the different services can be selected to import as a network.

  • PTEN IDs needed to be included in the analysis: ENSG00000171862 ENST00000371953 P60484 PTEN
  • The network titles can be edited and the different networks can be merged with the Cytoscape functions. The default style can be changed to the PSI-MI 2.5 Style for better interpretation.
Example of how the PTEN query results look like in PSICQUIC View. The results from IntAct are selected. This can be downloaded and imported as interaction networks in Cytoscape.

Results from PSICQUIC View

Example of how the PTEN query results look like in PSICQUIC View. The results from IntAct are selected. This can be downloaded and imported as interaction networks in Cytoscape.
Steps and results from the PTEN query in PSICQUIC core app of Cytoscape (version 3.8.2). The different PTEN IDs are selected in the PSICQUIC search tool bar in Step 1. In the new Cytoscape window, the interactions from the different databases are selected to import as a new network (Step 2). If different databases are selected, the resulting networks can be merged or displayed separately. Finally, the query visualization of the merged results show different networks with PTEN as a central node, in most of the networks.

Direct PSICQUIC query in Cytoscape

Steps and results from the PTEN query in PSICQUIC core app of Cytoscape (version 3.8.2). The different PTEN IDs are selected in the PSICQUIC search tool bar in Step 1. In the new Cytoscape window, the interactions from the different databases are selected to import as a new network (Step 2). If different databases are selected, the resulting networks can be merged or displayed separately. Finally, the query visualization of the merged results show different networks with PTEN as a central node, in most of the networks.

RESULTS

In each Cytoscape network, a central node is displayed representing PTEN. This is based on different IDs that are being used for PTEN, in the different resources, i.e. the UniProt ID is used by IntAct for protein:protein binding, the Ensembl gene ID is used by the University College London (UCL) team for annotating mRNA bound by miRNAs, whereas the Ensembl transcript ID is used by the IntAct team for the mRNA bound by microRNAs.

At this point, the query shows no TFs that regulate the expression of PTEN but several miRNAs were identified. Although in the Genome Browser of the University of California Santa Cruz (UCSC Genome Browser) the regulatory data suggests there are over 2000 TF binding sites associated with PTEN, this data is not available in the PSICQUIC dataset. 

At the moment, the primary interaction type is available through PSICQUIC (e.g. physical association, proximity, ubiquitination reaction, etc.) but the effect on the target entity is not yet available; therefore it is not possible to tell the difference between protein:protein binding or TF:TG binding and relational events. Nevertheless, it is assumed that miR:PTEN binding data in PSICQUIC would provide the list of miRs that regulate PTEN expression.

ADDITIONAL ANALYSIS

One thing that can be done to add the causality relationship information into the Cytoscape network is to use the SIGNOR database. The interactions already present in the query network might be described in SIGNOR, and, if so, these annotations can be added in Cytoscape to provide directionality in the network.

In SIGNOR the query for PTEN can be specified for the species Homo sapiens, selecting all relationships for the human protein. More so, the search can be refined for a specific mechanism (binding, phosphorylation, transcriptional regulation, etc.).

Next, the relationships can be downloaded and displayed in Cytoscape. The .tsv file downloaded from SIGNOR can be imported directly as a network and the columns can be edited to define the information correctly in Cytoscape.

Finally, the generated network from SIGNOR can be merged with the previous result from PSICQUIC query and more analysis can be performed.

Example of how the PTEN query results look like in SIGNOR. The query can be refined with the search tools (Step 1) and the interactions can be downloaded (Step 2) and imported to Cytoscape

SIGNOR query for PTEN interactions

Example of how the PTEN query results look like in SIGNOR. The query can be refined with the search tools (Step 1) and the interactions can be downloaded (Step 2) and imported to Cytoscape
Use Case 2
Main - Use cases
Use Case 3