e-Science Tools For The Genomic Scale Characterisation Of Bacterial Secreted Proteins

1
e-Science Tools For The Genomic Scale
Characterisation Of Bacterial Secreted Proteins

• Tracy Craddock, Phillip Lord, Colin Harwood and
  Anil Wipat

Newcastle University
2
Outline

• Computational challenges of bioinformatics
• Secretion in Bacillus
• Classification and analysis workflows
• Results and discussion

3
Computational Challenges of Bioinformatics

• New requirements from bioinformatics
• 3 major problems
• Heterogeneity
• Distribution
• Autonomy
• Experiments - series of workflows

4
myGrid and Taverna
Freefluo
Freefluo Workflow engine to run workflows
Scufl Simple Conceptual Unified Flow
Language Taverna Writing, running workflows
examining results SOAPLAB Makes applications
available
5
Microbase

• Grid-based system for microbial genome comparison
  and analysis
• Information repository (and execution
  environment)
• Pre-computed data

6
Outline

• Computational challenges of bioinformatics
• Secretion in Bacillus
• Classification and analysis workflows
• Results and discussion

7
Secretion in Bacillus

• Predict characteristics behavior of bacteria
• Identify secreted proteins
• Bacillus species diverse behaviour
• Soil inhabitants
• Harmful bacteria

8
Importance of Secretion

• Mechanism of interaction with environment
• Reveal capabilities of an organism
• Pathogens are of great interest

9
Secretory Proteins
Signal Peptide
Cytoplasm
Membrane
Cell Wall
Medium
Lipoprotein
Transmembrane
Cell wall binding
LPXTG
10
Outline

• Computational challenges of bioinformatics
• Secretion in Bacillus
• Classification and analysis workflows
• Results and discussion

11
Bioinformatic Tools
Signalp
Signal Peptide
Cytoplasm
Membrane
TMHMM tmap MEMSAT
LipoP
Cell Wall
ps_scan
Medium
Lipoprotein
Transmembrane
Cell wall binding
LPXTG
12
Classification Workflow
13
Process of Analysis
Putative secreted proteins
Protein families
Functional classification
Relations
14
Analysis Workflow
15
Architecture

• Custom-designed database

• Provenance tracking

• Analysis computationally intensive

• Architecture differs from other systems

16
Web Portal
17
Outline

• Computational challenges of bioinformatics
• Secretion in Bacillus
• Classification and analysis workflows
• Results and discussion

18
Classification Results
19
Functions of the Clusters
Number of families
20
Biologists Outlook

• Results available for subsequent analysis
• Data and results are of great interest

21
eScientists Outlook

• Microbase simplified data analysis
• But
• Autonomy - most services provided originally by
  external parties
• Licensing limits exposure of services
• Distribution - difficulty came from the
  relatively large datasets

22
Future Enhancements

• Use notification to automatically analyse
  recently annotated genomes
• Migrate workflows to a remote enclosed
  environment?

23
Acknowledgments

• Phillip Lord
• Colin Harwood
• Anil Wipat
• myGrid
• Carole Goble
• Tom Oinn
• and the rest of the myGrid team

• Microbase
• Yudong Sun
• Anil Wipat
• Matthew Pocock
• Pete A. Lee
• Paul Watson
• Keith Flanagan
• James T. Worthington