The GenomeMine: The need for improved genomic metadata capture and exchange

1
The GenomeMine The need for improved genomic
metadata capture and exchange

• Dawn Field
• Oxford Centre for Ecology and Hydrology

2
Deluge of Data
3
The GenomeMine
4
The Genome

• The H. influenzae genome was published in 1995
  and was the first bacterial genome to be
  completed. This genome is notable because it was
  the first bacterial genome sequenced.

5
NCBI Taxonomy

• The NBCI RefSeq number for this genome is
  NC_000907 and the taxid for this species is
  71421. This species belongs to the division of
  proteobacteria (gamma subdivision).

6
General Genomic Features

• The genome of index strain RD contains a single
  circular chromosome of 1,830,137 bp in size with
  an average GC content of 38. Strain RD has 1743
  ORFs, 6 rRNA operons, 54 tRNAs. Strain RD has 0
  plasmids, 0 mega-plasmids, and 0 bacteriophage.

7
Morphology and Growth

• Haemophilus influenzae is a non-motile,
  gram-negative, rod shaped bacterium. Optimal
  growth temperature is 37 degrees and doubling
  time in culture is 26 minutes.

8
Interactions and Ecology

• H. influenzae is a obligate commensal with the
  ability to cause disease including menigitis and
  otitis media. The primary habitat of this species
  is the human nasopharyx. This bacterium is
  faculatively anaerobic and uses organic matter as
  a source of carbon and organic matter as a source
  of energy.

9
Cataloguing our Complete Genome Collection

• Proposal Field D, Hughes J (2005). Cataloguing
  our current genome collection. Microbiology. 151,
  1016-9
• Hughes J Field D (2005) Ecological Perspectives
  on our complete genome collection Ecology
  Letters (in press).
• Workshop Cataloguing our current genome
  collection Sept 7-9, 2005 Cambridge, UK NIEeS
• Website Cataloguing our current genome
  collection (NERC International Opportunities
  Fund Award NE/3521773/1)

10
Cataloguing our Complete Genome Collection

• Proposal Field D, Hughes J (2005). Cataloguing
  our current genome collection. Microbiology. 151,
  1016-9
• More information describing our complete genome
  collection valuable (prokaryotes, eukaryotes,
  viruses, plasmids, and organelles)
• Complete set of information not readily
  available distributed widely among different
  sources
• A new standard could capture minimal information
  about a genome sequence
• developed within the auspices of an international
  working group
• E-Science implementation

11
MIGS Genome Reports

• Origin of Strain and Availability
• Geographical Origin and Environmental Context
• Ecology
• Taxonomy
• Annotation Process
• Genomic Features
• Publication and Electronic Resources
• Contact Information

12
MIGS Genome Reports

• Origin of Strain and Availability
• At a minimum this section should provide enough
  information for a user to easily request a stock
  sample of this genome (culture collection ID).
• Ideally, this section should contain a
  description of the salient features of the index
  strain (time in culture, risk of extinction, why
  it was selected, notable phenotypes)

13
MIGS Genome Reports

• Geographical Origin and Environmental Context
• At a minimum this section should provide enough
  information for a user to be able to re-sample
  the location (under similar environmental
  conditions) from which the genome was taken using
  the same isolation methodologies.
• Ideally, it should provide a range of information
  about the environment that an organism was taken
  from at the time of sampling.

14
MIGS Genome Reports

• Ecology
• At a minimum this section should provide
  information on the relationship between this
  genome and other genomes.
• Ideally, this section will allow the collection
  of a rich set of ecological information for each
  genome.

15
MIGS Genome Reports

• Taxonomy
• At a minimum this section should provide enough
  information for a genome to be uniquely and
  unambiguous identified by name in the literature.
  
• Ideally, it should provide the most update to
  date understanding of an organisms taxonomy
  possible.

16
MIGS Genome Reports

• The Annotation Process
• At a minimum this section could be optional.
• Ideally, it will give an overview of the
  annotation process, so that the same level of
  annotation could be regenerated from scratch.

17
MIGS Genome Reports

• Genomic Features
• At a minimum this section could be optional.
• Ideally, it should confirm a variety of genomic
  features that can not be calculated from the
  genome sequence.

18
MIGS Genome Reports

• Related Publications and Electronic Databases
• At a minimum this section should provide enough
  information to find the genome in at least one
  electronic resource (primary genome publication,
  international sequence database (e.g. Genbank)).
  
• Ideally it would provide sufficient information
  about all relevant resources containing
  information on this genome.

19
MIGS Genome Reports

• Contact
• At a minimum this section should contain the
  name and email of the person proposing to
  sequence the genome or the corresponding author
  on the primary genome paper.
• Ideally, it will contain contact details for
  other people with relevant areas of expertise
  including the person who submitted the genome
  (e.g. the bioinformaticians who did the genome
  annotation, the ecologist who isolated the
  organism, the taxonomist who is an expert on this
  group, etc).

20
Challenges

• Defining the standard
• Collecting the data
• Fields can be calculated in a variety of ways
  separate curated and calculated fields
• We dont know enough about many of these genomes
  with respect to lifestyle
• Relationships between genomes
• Completeness of data

21
Future uses

• Research tool
• Explore relationships between features
• Compare outputs of different methodologies
• Look-up Reference
• Network of researchers
• Teaching tool

22
Minimal Requirements

• must describe all genomes (eukaryotes,
  prokaryotes, plasmids, viruses, and organelles)
• must be easy to enter descriptions into the
  genome catalog
• possible to update descriptions / specification
• descriptions must be comparable across genomes
• descriptions easy to integrate with other
  resources
• each description must be associated with unique
  and permanent identifiers
• all information collected must be freely
  available

23
Discussion Document

• Table of Contents
• Executive summary
• The need for a new genomic standard
• The process of consensus building
• The specification checklist
• The specification draft version available
• Issues and challenges
• Considerations
• Case Studies
• Implementation options
• Genomic metadata exchange

24
Building Consensus

• Identification of the need for a standard
• The formation of a working group (community)
• Selection of case studies
• Development of the specification checklist,
  specification attributes, definitions, and
  allowed terms
• Development of a suitable implementation
• Development of a repository to store, view, and
  distribute annotations
• Final annotation of genomes to compliant format
  and submission to the repository.

25
Open call for Case Studies

• Contributing a first-pass description of your
  genome(s) will help provide input into the design
  of the specification
• Both descriptions and comments are collected from
  each contributor and these lead to the refinement
  of the specification

26
Call for contributed data sets

• GenomeMine re-designed to be a community archive
• Supplementary information for publications
• Archive of published data sets
• Selected data from other databases
• Each dataset captured in GnoME format
• Credit can be ascribed to author(s) (provenance)
• Data can be regenerated (variable definitions)

27
Summary

• The GenomeMine contains information about
  genomes data comes from a variety of sources
  aim is to serve as a community archive
  http//www.genomics.ceh.ac.uk/GMINE/
• From the difficulties associated with collecting
  curated data, the Cataloguing our complete
  genome collection project launched
• Open call for case studies (MIGS project) to
  describe genomes with curated information
• Open call for contributed data sets containing
  describing genomes to the GenomeMine using GnoME
  (curated, calculated and extracted variables
  which are defined)

28
Acknowledgements

• Genomemine
• Jennifer Hughes
• Tanya Gray
• Milo Thurston
• Gareth Wilson
• Adrian Tett
• Paul Swift
• Chimdi Ekeke

• Towards and new genomic standard
• Tatiana Tatusova, Jen Hughes, James
• Cole, Tanya Gray, Robert Feldman,
• Naomi Forrester, Dan Haft, Andy
• Lilley, Nick Mann, Victor Markowitz,
• Norman Morrison, Julian Parkhill, Lita
• Proctor, Jeremy Selengut, Susanna-Assunta
• Sansone, Paul Swift, Adrian Tett, Nick
• Thomson, Sarah Turner, Gareth Wilson,
• Anil Wipat
• and Everyone at this workshop

29
Genomic Metadata Exchange Format (GnoME)