1. Canadian Bioinformatics Workshops

2. Cold Spring Harbor Laboratory
In collaboration with
Cold Spring Harbor Laboratory
&
New York Genome Center

3. Module #: Title of Module3

4. Module 3 Lab Cytoscape & Enrichment Map
Jüri Reimand
High-throughput Biology: From Sequence to Networks
April 27-May 3, 2015

5. Network Analysis Workflow
A specific example of this workflow:
Cline, et al. “Integration of biological networks and gene expression data using Cytoscape”, Nature Protocols, 2, (2007).
TODO:
What: Schedules first part of talk
How
Bridge: So this is all about networks....

6. Pathways vs Networks - Detailed, high-confidence consensus
- Biochemical reactions
- Small-scale, fewer genes
- Concentrated from decades of literature
- Simplified cellular logic, noisy
- Abstractions: directed, undirected
- Large-scale, genome-wide
- Constructed from omics data integration

7. Networks Represent relationships of biological molecules
Physical, regulatory, genetic, functional interactions
Useful for discovering relationships in large data sets
Better than tables in Excel
Visualize multiple data types together
Discover interesting patterns
Network analysis
Finding sub-networks with certain properties (densely connected, co-expressed, frequently mutated, clinical characteristics)
Finding paths between nodes (or other network “motifs”)
Finding central nodes in network topology (“hub” genes)

8. Network basics (i): nodes and edges
A simple mapping
one molecule - node
one interaction - edge
A more realistic mapping
Cell localization
cell cycle
cell type
taxonomy
Physiologically relevant
Edges - diverse relationships
Critical: what do nodes and edges mean?
Edges and nodes : a network is composed of 2 elements : edges and nodes. We can have 2 types of network when we speak about the output of a pathway enrichment analysis.
A gene gene network or a pathway network. Node (circle) are going to represent gene and are going to be related by edges (lines) .
We also can have a pathway network : nodes (circle) are representing a gene-set and edge represents the number of overlapping genes

9. Network basics (i): nodes and edges
Node (molecule)
Gene
Protein
Transcript
Drug
MicroRNA …
A simple mapping
one molecule - node
one interaction - edge
A more realistic mapping
Cell localization
cell cycle
cell type
taxonomy
Physiologically relevant
Edges - diverse relationships
Critical: what do nodes and edges mean?
Edge (interaction)
Genetic interaction
Physical protein interaction
Co-expression
Signaling interaction
Metabolic reaction
DNA-binding
Directed or undirected?
Edges and nodes : a network is composed of 2 elements : edges and nodes. We can have 2 types of network when we speak about the output of a pathway enrichment analysis.
A gene gene network or a pathway network. Node (circle) are going to represent gene and are going to be related by edges (lines) .
We also can have a pathway network : nodes (circle) are representing a gene-set and edge represents the number of overlapping genes

10. Network Representations

11. Network basics (ii) - layout and visual attributes
Use layout to
interpret network
Local relationships
Guilt-by-association
Dense clusters
Global relationships
Visualise multiple
types of data

12. Automatic network layout

13. Force-directed network layout
Force-directed: nodes repel and edges pull
Good for up to 500 nodes
Bigger networks give hairballs - reduce number of edges
Advice: try force directed first, or hierarchical for tree-like networks
Tips for better looking networks
Manually adjust layout
Load network into a drawing program (e.g. Illustrator) and adjust labels

14. force directed layout 1 2 3 4

15. Dealing with ‘hairballs’: zoom or filter
Focus
PKC Cell Wall Integrity

16. Visual Features Node and edge attributes Visual attributes
Attribute color
Arrow shape
Visual Features
Node and edge attributes
Represent properties of genes, interactions
Text (string), integer, float, Boolean, list
Visual attributes
Node, edge visual properties
Colour, shape, size, borders, opacity...
Edge shape
Node shape

17. What Have We Learned?
Networks help discover relationships in large data sets
Networks help integrate several datasets or types
Important to understand meaning of nodes and edges
Avoid hairballs by focusing analysis
Automatic layout is required to visualize networks
Visual attributes enable multiple types of data to be shown at once – useful to see their relationships 7
Part II

18. Cytoscape - Network Visualization
Cytoscape is
an open source software platform
for visualizing complex networks and integrating these with any type of attribute data.
a lot of apps are available for various kinds of problem domains, including bioinformatics, social network analysis, and semantic web.
Network visualization can be done using the Cytoscape software. You can create and visualize networks using Cytoscape.

19. Manipulate Networks
Filter/Query
Interaction Database Search
Automatic Layout

20. The Cytoscape App Store
Pathway analysis
Gene expression analysis
Complex detection
Literature mining
Network motif search
Pathway comparison

21. Introduction to Cytoscape (3.1.0)
save your session
save image
Results panel
Control panel
And finally these next slides represent a quick Cytoscape tutorial.
Cytoscape has 3 major panels : control panel, data panel and results panel.
You can save your session as a .cys file and reopen it later
Table panel

22. Navigate through the network (3.1.0)
move the blue square to navigate through the network
When you want to navigate through the network, you can use move the blue square in the network window of the control panel 7
Part II

23. Try different Cytoscape layouts (3.1.0)
Circular Layout
If you want to redo your layout, you can go to layout in the menu bar and choose the layout of your choice 7
Part II

24. yFiles Circular

25. yFiles Organic

26. Change visual Features (3.1.0)
STYLE
Colors
Shapes
Sizes
Brushes
Transparency …
You can play with visual features using Vizmapper in the control panel window 7
Part II

27. Visual Style Load “Your Favorite Network”
File > Import > Network > File …

28. Visual Style Load “Your Favorite Expression” Dataset
File > Import > Table > File …

29. Visual Style
Map expression values to node colours using a continuous mapper

30. Visual Style
Expression data mapped
to node colours

31. Fine-tuning network layout
Move, zoom/pan, rotate, scale, align

32. Create Subnetwork

33. Create Subnetwork

34. Select nodes with at least 20 interactions
Network Filtering
Select nodes with at least 20 interactions

35. Interaction Database Search
Query the BioGrid database for interactions of the cancer gene KRAS

36. What have we learned
Cytoscape is a useful, free software tool for network visualisation and analysis
Provides basic network features
Automated layouts
Mapping of genomic attributes to visual attributes
Apps are available to extend functionality to diverse analyses

37. Enrichment Map
A Cytoscape app to visualize and interpret results of pathway enrichment analysis
Enriched pathways are visualized as networks
Edges connect pathways with many shared genes
pathway network
node
(pathway)
edge: # of overlapping genes

38. GO.id GO.name p.value covercover.rat Deg.mdn Deg.iqr
GO: taxis E
GO: chemotaxis E
GO: adaptive immune response based on somatic recombination E
GO: adaptive immune response E
GO: leukocyte mediated immunity
GO: B cell mediated immunity
GO: myeloid cell differentiation
GO: immune effector process
GO: regulation of phagocytosis
GO: positive regulation of phagocytosis
GO: lymphocyte mediated immunity
GO: growth factor binding
GO: protein polymerization
GO: endoplasmic reticulum membrane
GO: immunoglobulin mediated immune response
GO: heart development
GO: response to bacterium
GO: regulation of endocytosis
GO: acute inflammatory response
GO: positive regulation of endocytosis
GO: myeloid leukocyte activation
GO: amino acid biosynthetic process
GO: regulation of inflammatory response
GO: activation of immune response
GO: positive regulation of immune system process
GO: positive regulation of immune response
GO: antigen processing and presentation
GO: regulation of immune system process
GO: regulation of immune response
GO: negative regulation of enzyme activity
GO: phagocytosis
GO: myeloid leukocyte differentiation
GO: humoral immune response
GO: lymphocyte activation
GO: leukocyte chemotaxis
GO: negative regulation of protein kinase activity
GO: negative regulation of transferase activity
GO: transforming growth factor beta receptor signaling pathw
GO: insulin-like growth factor binding
GO: T cell activation
GO: humoral immune response mediated by circulating immunogl
GO: cytosolic ribosome (sensu Eukaryota)

40. Creating an Enrichment Map with data from g:Profiler
Go to g:Profiler website - .
Select and copy all genes in the tutorial file MCF7_24hr_topgenes.txt in the Query box
In Options, check Significant only, No electronic GO annotations
Set the Output type to Generic Enrichment Map (TAB)
Show advanced options
Set Max size of functional category to 1000 and Min size to 5.
Set Min size of gene list and functional category overlap Q&T to 2.
Set Significance threshold to Benjamini-Hochberg FDR .
Choose GO biological process, molecular function, KEGG and Reactome from the color legend.
Download g:Profiler data as gmt: name
Click on g:Profile! to run the analysis
Download the result file: Download data in Generic Enrichment Map (GEM) format

41. g:Profiler files for Enrichment Map
Two files are required:
Tabular text file with enriched processes&pathways
GMT file with all processes&pathways and associated genes A. B.

42. TAB file – Enrichment Map
Process ID
Process name
P-value, FDR
Phenotype
positive 1; negative -1
Genes common to input list and process/pathway

43. GMT file – Enrichment Map
Process ID
Process name
Gene list

44. Getting the Enrichment Map app
A. Cytoscape App Store
B. In Cytoscape software
Apps > App manager > Search > Install

45. Constructing an Enrichment Map
A. Set input file format to “Generic”
(format of g:Profiler results) A.
B. Select GMT file from file system
(this file contains gene lists of pathways) B. C.
C. Select enrichments file from file system D.
D. Set equal P-value and Q-value cutoffs
(because g:Profiler only provides corrected P-values=Q-values)
Click Build !

46. Enrichment Map – a network of pathways
Nodes – pathways, processes, functions
Edges – between pathways with many shared genes
Node size – genes in pathway
Node color – enrichment strength
Edge weight – genes shared

47. Enrichment Map – a network of pathways
What are major functional themes?
Apoptosis
Prolife-
ration
Mitochondrial
processes
Protein
degradation
Tissue development

48. Edge definition determines granularity
Edges between pathways defined by % of common genes
Increasing edge stringency reveals finer granularity of functional themes
Protein
degradation
Mitochondrial
processes
Apoptosis

49. Mapping attributes to network Node labels mapped to pathway IDs by default
Pathway names
Pathway IDs

50. Compare two pathway enrichment analyses g:Profiler analyses of MCF7_12hr_topgenes.txt and MCF7_24hr_topgenes.txt
1st dataset maps to node fillings
2nd dataset maps to node edges

51. Mapping further gene sets to Enrichment Map E. g
Mapping further gene sets to Enrichment Map E.g. which processes associate to known cancer genes?
CANCER_GENES
B. Select GMT file from file system
(this file contains gene lists used in original Enrichment Map)
B. Select another GMT file from file system
(this file contains gene lists mapped on top of original Enrichment Map)

52. Post-Analysis: New edges link further gene lists to existing pathways
Purple edges – pathway genes in this analysis
are significantly enriched in known cancer genes

53. What have we learned?
Enrichment map uses network visualization to summarize results of pathway enrichment analysis
Edges connect pathways with shared genes, and edge stringency determines granularity of maps
Two sets of pathways can be compared
Post-analysis links further gene sets to existing map

54. Tips for publishable Enrichment Maps
Manually curate clusters of connected pathways as “functional themes”
e.g. by picking most general process/pathway in the group
Check genes involved
Review small groups and singleton nodes
Many are probably part of larger groups – safe to remove
Some interesting singletons may be the only representatives!
Assign further visual attributes using your omics data
Export as PDF and use graphics software (AI) for finalizing
Sometimes less is more

55. Cytoscape Tips & Tricks
Network views
When you open a large network, no view is shown by default
To improve interactive performance, Cytoscape has the concept of “Levels of Detail”
Some visual attributes will only be apparent when you zoom in
The levels for various attributes can be changed in the preferences
To see what things will look like at full detail:
ViewShow Graphics Details

56. Cytoscape Tips & Tricks
Sessions
Sessions save pretty much everything:
Networks
Properties
Visual styles
Screen sizes
Saving a session on a large screen may require some resizing when opened on your laptop

57. Cytoscape Tips & Tricks
Memory
Cytoscape uses lots of it
Doesn’t like to let go of it
An occasional restart when working with large networks helps
Destroy views when you don’t need them
Java doesn’t have a good way to get the memory right at start
Since version 2.7, Cytoscape does a much better job at “guessing” good default memory sizes than previous versions

58. Cytoscape Tips & Tricks
CytoscapeConfiguration directory
In your user/home directory
Your defaults and any apps downloaded from the app store will go here
Sometimes, if things get really messed up, deleting (or renaming) this directory can give you a “clean slate”

59. We are on a Coffee Break & Networking Session