EGGSViz : Visualization and Exploration of Gene Clusters

1
EGGSViz Visualization and Exploration of Gene
Clusters

• Ankita Bhan
• Advisor Prof. Sun Kim
• Co-advisor Prof. Yves Brun
• Indiana University, Bloomington

2
Outline

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration Feature 1
• Exploration Feature 2
• Future Work

3
Background

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

• Functionally related genes co-evolve, probably
  due to selection pressure during evolution.
• This leads to conservation of gene clusters
  across genomes (especially in microbial genomes).

4
Motivation

• Microbial genomes contain an abundance of genes
  with conserved proximity.
• Genes with conserved proximity are often
  co-transcribed as operons, or co-regulated as
  part of a larger biochemical network.

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

5
Motivation

• Gene clusters-Definition
• Group of genes in microbial genomes with
  conserved proximity that are the possible
  candidates for being co-transcribed as operons,
  or co-regulated as part of a biochemical pathway.
  

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

6
EXAMPLE OF A GENE CLUSTER
7
THE FUNCTIONAL CATEGORY CODES FROM TIGR
8
Problem Description

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

• Predicting sets of families in genomes by
  interpreting a genome as a sequence of families
  modeling as a gene cluster.
• Visualizing the clusters with the multiple
  genomes on a single display window is challenging
  as the clusters are scattered.
• Adding genes from a new unknown genome to a known
  cluster of genes is challenging.

9
Features

• simultaneous visualization of all gene clusters
  on genome scale with zoom in/out features
• detailed view of individual cluster
• color coding scheme according to COG functional
  categories
• multiple sequence alignment of genes in a cluster
• selection of clusters by feeding in
  "genes-of-interest" by users
• adding a new genome and search for instances of
  clusters and saving the results of search

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

10
Features Illustration of how EGGSVIZ works

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

• Workflow1
• Main display of EGGSVIZ and broader view of
  cluster.
• Workflow2
• Detailed view of an individual cluster
• Workflow 3
• Further details of each gene in the Detail View
  window.
• Exploration feature 1
• Visualizing our genes of interest.
• Exploration feature 2
• Adding a new genome and saving the search.

11
Workflow 1

• Display of all clusters on genome scale on a
  single display screen.
• Divided dynamic zoom (from low to high
  resolution) of genomic regions.
• Displays the cluster number and size(number of
  genes ) in a particular cluster in the main
  window.
• Highlights an individual cluster and shows the
  annotation information for each and every gene
  and cluster.

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

12
DISPLAY WINDOW
13
OPEN CLUSTER FILE
14
CLUSTER FILE LOADED
15
GENOMES ZOOMED OUT
16
GENOMES ZOOMED IN
17
MAIN DISPLAY WITH CONNECT CLUSTERS OPTION
18
CLUSTER OF INTEREST HIGHLIGHTED
19
COLOR REVERSE OPTION
20
CONNECT GENES OPTION
21
Workflow 2

• How to choose a cluster?
• Detailed view of the individual cluster of our
  choice.
• Color Reverse and Disconnect genes options.

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration1
• Exploration2
• Future Work

22
HOW TO GO TO THE DETAIL VIEW WINDOW
23
THE DETAIL VIEW WINDOW FOR CLUSTER 1 CONTAINING
30 GENES
24
GENE SIZE ZOOMED IN TO A SCALE OF 15
25
GENE SIZE ZOOMED IN TO A SCALE OF 25
26
DETAILED VIEW WITH THE COLOR REVERSE OPTION
27
DETAIL VIEW WITH DISCONNECT GENES OPTION
28
DETAIL VIEW WITH THE GENE ANNOTATION
29
DETAIL VIEW WITH THE GENE ANNOTATION
30
Workflow 3

• On double clicking a gene redirection to sequence
  window.
• Sequence window retrieves sequences of genes
  related in the synteny.
• Clustalw button facilitates the multiple
  alignment of sequences.
• More features to be added to connect this
  information to web services.

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration 1
• Exploration 2
• Future Work

31
HIGHLIGTING A CLUSTER
32
DETAILED VIEW OF CLUSTER 14 IN COLOR REVERSE
33
HIGHLIGHTING AN INDIVIDUAL GENE
34
RETRIEVING SEQUENCES OF THE HIGHLIGHTED SET OF
GENES
35
MULTIPLE ALIGNMENT OF SEQUENCES
36
Exploration Feature 1

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration 1
• Exploration 2
• Future Work

• A feature to explore the genes of interest is
  provided.
• Input button on the main window would prompt a
  window asking our genes of interest.
• On submitting the query we would get the complete
  detail of those genes and the genes would be
  highlighted in the detail view window.

37
INPUT BUTTON ON THE MAIN DISPLAY WINDOW
38
FEEDING IN THE GENES OF OUR INTEREST
39
INFORMATION RETRIEVED ON USING THE SUBMIT BUTTON
40
CLUSTER-gt GENES INVOLVEDTABLE REDIRECTS TO
DETAIL VIEW WINDOW
41
REDIRECTED TO DETAIL VIEW WINDOW
42
GENES OF INTEREST HIGHLIGHTED
43
GENES OF INTEREST HIGHLIGHTED IN YELLOW
44
GENES OF INTEREST HIGHLIGHTED
45
Exploration Feature 2

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration 1
• Exploration 2
• Future Work

• Import an additional new Genome to the cluster on
  the detail view window.
• Connect that genome to the present cluster and
  predict clusters by connecting to a web server.
• A Hidden Markov Model is used to predict similar
  genes from the 4th genome.

46
SELECTING THE CLUSTER OF YOUR CHOICE
47
DETAILED VIEW OF THE HIGHLIGHTED CLUSTER
48
ON CLICKING THE BROWSE BUTTON
49
ON SELECTING THE NEW GENOME
50
DETAILS OF THE GENES OF THE NEW GENOME
51
Future Work

• Background
• Motivation
• Problem Description
• Features
• Workflow 1
• Workflow 2
• Workflow 3
• Exploration 1
• Exploration 2
• Future Work

• Saving the results of the clusters generated
  after adding the 4th genome for faster and
  efficient lookup.
• Extending the cluster prediction and
  visualization beyond 3 genomes.
• Analyzing the gene clusters Phylogenetically and
  visualizing the results.

52
Acknowledgements

• Special thanks to Justin Choi, Center of
  Genomics and Bioinformatics, Indiana University
• Prof. Sun Kim, School of Informatics, Indiana
  University
• Prof. Yves Brun, Department of Biology, Indiana
  University
• Kwangmin Choi, Youngik Yang,School of
  Informatics, Indiana University
• Pamela Bonner and the Brun Lab, Department of
  Biology, Indiana University
• Faculty and the staff , School of Informatics,
  Indiana University