IXodus a knowledge discovery process based on the SIMDAT-Pharma GRID technologies

1
IXodusa knowledge discovery process based on the
SIMDAT-Pharma GRID technologies

• Richard Kamuzinzi
• Université Libre de Bruxelles Bioinformatics
• June, 5 7th 2007
• World Wide Workflow GRID ASIA 2007
• Singapore

2
SIMDAT Facts

• EU Information Society Technologies (IST)
• GRID Project
• Duration 4 years
• Start date September 1st 2004
• 26 partners

3
Scope

• Product and Process Development (automobiles,
  aircraft, drugs, meteorological services) is
• Complex
• Involves several independent organizations at
  different locations
• Complexity management in one site is too
  expensive gt cost/risk sharing with partners gt
  GRID

4
Strategic objectives

• to test and enhance Data Grid technology for
  product development and production process
  design,
• to develop federated versions of problem-solving
  environments by leveraging enhanced Grid
  services,
• to exploit Data Grids as a basis for distributed
  knowledge discovery,
• to promote defacto standards for these enhanced
  Grid technologies across a range of disciplines
  and sectors as well as
• to raise awareness of the advantages of Data
  Grids in important industrial sectors

5
Project organization (SIMDAT-Pharma)
NEC, GSK, Inpharmatica, ULB, Fraunhofer
SCAI-Bio and UKA
6
IXodus The scientific problem

• Lyme disease significant source of human and
  animal pathology in temperate areas of the world
  (identified in 90s)
• Caused by the bite of a tick of genus IXodes,
  infected by the pathogen bacterium Borrelia
  burgdorferi
• the study of host-parasite interactions is an
  active research as 20 ticks have been found
  infected by the bacterium
• IXodus scientific protocol designed to deal with
  characterisations of genes expressed in the
  salivary gland of the tick IXodes ricinus at
  various stage of the host-parasite interaction
  process

7
IXodus Workflow design (1)

• From IXodus scientific protocol to IXodus
  workflow (WF) design, we identify 2 uses cases
• New cDNA sequences the workflow is daily
  feeded with a batch of nucleic sequences from the
  systematic sequencing of thousands of salivary
  gland cDNAs
• Databank update whenever a new version of
  relevant biological databank appears, the core
  workflow analysis is re-enacted to discover
  potentially new information

8
IXodus design (2) Use Case 1
9
IXodus design (3) Use Case 2
10
IXodus Implementation

• Workflow technology platform InforSenseTM KDE
• Implementation is tightly coupled with the
  deployment environment, which is mainly driven by
  2 kind of constraints
• GRID approach
• Semantic Web (SW) approach

11
IXodus implementation - The test-bed GRID
approach
Knowledge DB IXodus
12
IXodus implementation - The test-bed SW approach
Semantic Broker
Service advertising
13
IXodus implementation InforSense KDE The
complete Workflow
14
IXodus implementation InforSense KDE User
sequences gathering
15
IXodus implementation InforSense KDE
Management of sequences overlapping
16
IXodus implementation InforSense KDE Main
analysis flow (Bioinformatics tools)
17
IXodus implementation InforSense KDE Service
instance selection launching
18
IXodus - General benefits

• Workflow tool maturity design of complex WF to
  support demanding problem in a reasonable
  delivery-time is a reality (RWD vs. RAD)
• WF on GRID approach is really valuable and
  provides the confidence we need to front the
  data/services tsunami in Life sciences the
  good news is

19
IXodus - General benefits (2)
...thanks to WF technologies, the scientists no
more scares the vertiginous beast
(data/services explosion)
20
IXodus Remaining challenges

• B2A Grids we still need precise understanding of
  strategic benefits from both (win-win) side
• WF technologies need better distinction between
  abstract WF and operational WF
• How to decouple?
• Runtime service selection using the concept of
  rules?
• At design phase the designer would appreciate
  semantics approach to search for services
• From WF to Service
• Partial (?args) vs. Complete(?args)
• Different profiles of user
• From WF to UI
• At design phase need to define how WF actors
  interact with the whole system
• To leverage the WF log in order to generate
  textual information that would support scientific
  papers/notebooks writing (who, service_name,
  service_version, database_version, )

21
SIMDAT- Major outcomes to expect

• SIMDAT approach will provide state-of-the-art
  components
• To enable industry-strength environment for
  e-Science activities
• To support the academia/industry collaborations
  in RD activities (B2B B2A Grids)
• B2A Grids how the win-win model is precisely
  configured?
• To help build up virtual organisations that
  federate data, services and scientific expertise

22

• Thank you !

• Web http//www.simdat.org
• Contact richard.kamuzinzi_at_ulb.ac.be
• Acknowledgments
• co-author Robert Herzog, Université Libre de
  Bruxelles (ULB)
• Scientific expert Valérie Ledent, ULB
• Edmond Godfroid Bernard Couvreur Laboratory of
  Applied Genetics, ULB
• SIMDAT colleagues Joseph Mavor (ULB), Falk
  Zimmermann (NEC), Changtao Qu (NEC), Nabeel Azam
  (InforSense), Moustapha Ghanem (InforSense), Kai
  Kumpf (SCAI-Bio)