MUGEN Network and associated database(s)

1
MUGEN Network and associated database(s)

• Helen Parkinson
• European Bioinformatics Institute, Hinxton, UK
• Mouse Phenome Database Meeting
• Barcelona, 25 February 2006

2
The MUGEN Network (NOE)

• Functional genomics approaches in animal models
  to study human immunological disease
• 25 partners
• Technical partners Arexis AB (Database),
  Genopolis (Analysis pipeline), EBI (microarray
  data, data standards, curation, Alexander
  Fleming, web portal), BioVista (text mining)
• Large microarray component, users surveyed for
  their abilities and needs for informatics support
• Database developed for integrated genotype and
  phenotype data (230 mouse strains, mostly KO/KI
  within MUGEN)

3
Database WP Technical objectives

• Establish an oracle-based MUGEN data environment
  that links all groups and facilitates exchange of
  data
• To develop the system to accommodate the changing
  needs of all partners
• To establish within the environment relational
  database for phenotypes, pathways and loci

4
Arexis MUGEN Database

• Java, Postgres
• Status data acquisition in user testing
• Data mining tools being specified
• Data types, genotypes, phenotypes
• No cv determined yet for describing phenotypes
  within MUGEN
• Microarray and sequence data not held locally
  links to ArrayExpress, Genopolis, and Ensembl
  (mouse, human)
• Annotation standards in place for microarray data
  for MUGEN

5
What would we like for MUGEN

• Tools to provide access to the phenome data
  generated within MUGEN
• Common vocabulary/ontology for describing mouse
  phenome data that works for the MUGEN data
• Tools for accessing/importing remote phenome data
• A community reporting std that we can conform to
  for the phenotype data

6
FUGO

• Functional genomics investigation ontology
  experiments, technologies proteomics,
  transcriptomics, metabolomics etc, domains
  toxicology, model organism biology, core plus
  community specific extensions.
• Upper level ontology v1 in place now being
  extended with general concepts not species
  specific
• Process, object, characteristic mapping
  underway by communities
• Built in Protégé-OWL
• OBO application ontology (?)
• fugo.sourceforge.net fugo-devel_at_lists.sourceforge.
  net

7
Lessons learned from dealing with standards at EBI

• Social engineering is the first and biggest
  hurdle
• Transparency is good open mailing lists, common
  documents, outreach
• Content, syntax, format important applications
  are key, simple is good
• Rapid prototyping of ontologies, models, formats
  and apps is useful for community buy-in
• Dont forget the users we (MGED) have over
  engineered things in the past

8
Thanks to

• Thomas Svensson, Tobias Terstad, Mikael Lantz
  Arexis AB, Sweden
• Werner Mueller, MUGEN/Eumorphia
• Vassilis Aidinis, George Kollias - Alexander
  Fleming, Greece
• Microarray Informatics Team, EBI
• Genopolis Team - Milan
• Ensembl Team EBI
• Eumorphia
• Funding (FELICS, MUGEN, NHGRI, NCRI, NIH)