myGrid overview

1

• myGrid - overview
• myGrid is an extensible open platform for
  e-Science data tools interoperability built
  using existing Web services and Grid technologies
  that supports
• In silico experiments based on process flows
• Data provenance and resource change management
  based on notification and process flow evolution
• Explicit capture of the e-scientists knowledge
• Personalised views over repositories,
  personalised process flows, and personal data
  sets
• The myGrid toolkit can be configured for specific
  applications, building on the experience of the
  consortium in user requirements capture and
  community-based tools

Exploring Williams-Beuren Syndrome Using
myGrid Robert Stevensa, Hannah J. Tipneyb, Chris
Wroea, Tom Oinnc, Martin Sengerc, Phillip Lorda,
Carole Goblea, Andy Brassa, May Tassabehjib a
Department of Computer Science, University of
Manchester, Oxford Road, Manchester, United
Kingdom, M13 9PL b University of Manchester
Academic Unit of Medical Genetics, St Marys
Hospital, Hathersage Road, United Kingdom M13
0JH c European Bioinformatics Institute, Wellcome
Trust Genome Campus, Hinxton, Cambridge, United
Kingdom CB10 1SD

• The Biological Problem
• Williams-Bueren Syndrome (WBS) is a rare,
  sporadically occurring disorder characterised by
  a unique set of physical and behavioural
  features. WBS is caused by a deletion located in
  chromosome band 7q11.23, in a region flanked by
  highly repetitive regions containing both genes
  and pseudogenes.
• Most WBS inidividuals have a deletion of about
  1.5Mb, encompassing 24 genes (see right), but a
  smaller region containing the genes critical to
  the WBS phenotype has been identified. This
  smaller reigion is known as the WBS Critial
  Region (WBSCR).
• The WBSCR has not yet been fully mapped,
  primarily because of its complex and repetitive
  nature. The gaps in the WBSCR may harbour
  important genes and associated regulatory
  elements. The purpose of this myGrid application
  is to help produce the complete, comprehensive
  and robust map of the WBS region that is vital if
  we are to fully understand the pathology of WBS.

• The e-Science Process
• In silico experiments necessitate the virtual
  organisation of people, data, tools and machines.
  The scientific process also necessitates an
  awareness of the experience base, both of
  personal data as well as the wider context of
  work. The management of all these data and the
  co-ordination of resources to manage such virtual
  organisations and the data surrounding them needs
  significant computational infrastructure support.
  
• myGrid, middleware for the Semantic Grid, enables
  biologists to perform and manage in silico
  experiments, then explore and exploit the results
  of their experiments.

• The Bioinformatics Experiment
• We have developed a workflow language, Scufl, a
  workflow development environment, Taverna, and a
  workflow enactment engine, Freefluo, that allow
  biologists and bioinformaticians to represent an
  experiment design explicitly without the
  complication of writing a complex bespoke
  application .
• As well as orchestrating the execution of the
  workflows service components, the enactor can
  also generate provenance information annotations
  in RDF under the control of user-defined
  annotation templates.
• The diagram (right) shows a schematic
  representation of the first workflow created to
  explore gap regions within the WSBCR. This
  workflow takes the last verified piece of
  sequence (lt 3000 bp) in the contig flanking a
  gapped region and produces a shortlist of
  sequences which may extend the contig into the
  gap region.

• Result Co-ordination
• Each run of a series of experiments produces a
  large number of data files data are produced for
  each service and for each input, multiple outputs
  can be generated.
• To validate results, a biologist needs to be able
  to trace back through data from each part of the
  analysis.
• In addition, a biologist needs to look back
  through a history of experiments on a particular
  topic look at experiments on a different topic
  look at colleagues experiments and also view
  experiment data holdings in a variety of views
  suited to the current needs.
• in myGrid, this personalisation comes from
  co-ordinating these complex, inter-related data
  holdings acording to the myGrid information model
  through decoration with RDF and LSID references.
• The data graph produed by Freefluo (below) is the
  counterpart to the workflow graph (left), where
  the data are the nodes of the graph and the arcs
  the processes that produced those data.
• myGrid uses Haystack to enable the biologist to
  view this graph of results and follow the RDF
  links between results.

• Outcomes
• Performing such results manually through Web
  based resources can take at least two days of
  tedious, error prone cutting and pasting between
  a host of Web pages.
• These myGrid workflows take about one hour to run
  and produce a collection of co-ordianted results
  that facilitate analysis and management of an
  experiment's data holdings.
• This increase in efficiency in performing an
  analysis is coupled with the ability to easily
  replicate the expeirmental protocol and gives a
  systematic managment of results.
• The generic results co-ordination system enables
  a biologist to create an experience base of
  experimental techniques, data holdings and other
  organisational information that facilitates
  personalisation of e-Science.
• The abstract, declarative nature of the workflows
  means that creation of an analysis provides an
  alternative to the writing of bespoke software.
• Using myGrid in this way has extended the genetic
  map into the WBS Critical Region.