Network visualisation

Results of active module identification (AMI) and random walk with restart (RWR) approaches can be visualised using Cytoscape [1].

This page contains a guideline, step by step, to create network visualisation of results such as those presented in this documentation.

Active module identification results visualisation

Tip

To run an Active Module Identification with data retrieved by queries, see Use-case 1 and with data provided by user, see Use-case 2.

The Fig. 6 is an example of AMI results visualisation.

Fig. 6 : Use-case 1 AMI analysis visualisation (from Fig. 19). We use Cytoscape [1] to create network visualisation and Omics Visualizer [2] to add overlap results to active modules.

Step by step guidelines

To visualise the AMI results using network representation, use the following steps:

1. Import files

Fig. 7 : Step 1 - Import files

• Import Network from File: DOMINO_*_activeModulesNetwork.txt

• Import Table from File: DOMINO_*_activeModules.txt (Import Data as Node Table Columns)

2. Add donuts

Fig. 8 : Step 2 - Add donuts using OmicsVisualizer [2] app

1. Omics Visualizer [2] table from File: DOMINO_*_overlapAMresults4Cytoscape.txt

2. Manage table connections: link network node table with right identifiers

   • Network: sharedname

   • Table: geneSymbol

3. Donut Chart Visualisation: Add overlap results using donuts (Use termTitle)

3. Network style

Fig. 9 : Step 3 - Apply style that you want

• Border Width = 5.0

• Fill Color = ActiveGenes

  • True #CCCCCC

  • False #FFFFFF

• Label Font Size = 20

• Shape = Ellipse

• Size = 50

• Lock node width and height

4. Active module selection

Fig. 10 : Step 4 - Select identified active module with a significant overlap

1. Filter: Select nodes with overlapSignificant = True

2. New Network: From Selected Nodes, All Edges

Tip

You can select modules that you are interested in directly (Ctrl + mouse drag) then create a new network from selected nodes (step 2 above).

5. Create legends

Fig. 11 : Step 5 - Add legend using Omics Visualizer [2]

Random walk with restart results visualisation

Tip

To perform a RWR with data retrieved by queries, see Use-case 1 and with data provided by user, see Use-case 2.

The Fig. 12 is an example of RWR results visualisation.

Fig. 12 : Use-case 1 RWR analysis visualisation (from Fig. 20). We use Cytoscape [1] to create network visualisation.

Step by step guidelines

To visualise the RWR results using network representation, use the following steps:

1. Import files

Fig. 13 : Step 1 - Import files

• Import Network from File: UseCase1_RWR_network.sif

• Import Table from File: multiplex_1.tsv and multiplex_2.tsv

• Import Table from File: seeds.4Cytoscape

  • Change column names: node for column 1 and seed for column 2

Tip

How create the seeds.4Cytoscape file ?

awk -F"\t" 'NR==FNR{a[$1]; next} {if($2 in a){print $2"\tTrue"}else{print $2"\tFalse"}}' seeds.txt multiplex_1.tsv > seeds.4Cytoscape

• Import Table from File: diseasesDescription.txt

  • Change column names: node for column 1, pathways for column 2 and score for column 3

Tip

How create the diseasesDescription.txt file ?

awk -F"\t" 'NR==FNR{a[$1]=$2; next} {if($1 in a){print $1"\t"$2"\t"a[$1]}}' RWR_top20.txt ../../OutputOverlapResults/WP_RareDiseases_request_2022_09_07.gmt > diseasesDescription.txt

2. Management of nodes table

Fig. 14 : Step 2 - Create two new columns

• Create two new columns named label as string and keep as boolean in the node table

Fig. 15 : Step 2 - Select nodes using Filter and fill node table

• Filter: Select genes nodes (multiplex is 1)

  • Fill label column with =$name and apply to selected nodes

  • Fill keep column with =$seed and apply to selected nodes

  • Sort by score (decrease) and select the 30th first genes that are not a seed (selected nodes from selected rows)

  • Fill keep column with =True and apply to selected nodes

• Filter: Select are disease pathways nodes (multiplex is 2)

  • Sort by score (decrease) and select the 5th first rare disease pathways

  • Fill keep column with =True

  • Fill label column with =$pathways and apply to selected nodes

3. Create new network

Fig. 16 : Step 3 - Select nodes using Filter, create a new network and remove duplicate edges

• Filter: Select nodes with keep = True

• New Network: From Selected Nodes, All Edges

• Edit and Remove Duplicate Edges

4. Network style

Fig. 17 : Step 4 - Apply style

• Change the style of nodes

Table 1 - Network Style

	All Nodes	Disease Nodes
Border Width	5.0	5.0
Fill Color	Column seed True: CCCCCC False: FFFFFF	DD3497
Label Front Size	20	50
Shape	Ellipse	Triangle
Lock node width & height	True	True
Size	50	100
Label	label	label

• Change network layout (here we used yFiles Organic Layout)

References

[1] (1,2,3)

Shannon P, Markiel A, Ozier O et al.. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome research. 2003.

[2] (1,2,3,4)

Legeay M, Doncheva NT, Morris JH et al.. Visualize omics data on networks with Omics Visualizer, a Cytoscape App. F1000Research. 2020.