Sharing Genomic Data and Annotations using GFF3 format

1
Sharing Genomic Data and Annotations using GFF3
format
Dina Sulakhe and Natalia Maltsev Bioinformatics
Group MCS, Argonne National Laboratory Computation
Institute University of Chicago
2
What we are going to talk about?

• GFF3 overview
• GFF3 Standards
• For Sharing Annotations
• For cross-referencing the data
• Extending GFF3
• Adding annotations from public databases
• Adding users annotations
• Sharing and exchanging annotations using
  Web-services
• GFF3 genomes repository at Argonne
• Downloads
• Web-services

3
GFF3 overview (Lincoln Stein, 2004)

• A tab-delimited flat file representation of
  genomic features
• GFF3 format
• provides a mechanism for representing of
  hierarchical grouping of genomic features and
  sub-features
• separates the ideas of group membership and
  feature name/id
• Enforces the use of controlled vocabularies by
  imposing constraints on the definitions of
  genomic features
• allows a single feature (e.g. an exon) to belong
  to more than one group at a time.
• provides an explicit convention for pair wise
  alignments
• provides an explicit convention for features that
  occupy disjoint regions

4
An Example
5
PUMA2/GNARE Systems

• PUMA2 (http//compbio.mcs.anl.gov/puma2) is an
  Interactive Integrated Environment for
  High-throughput Genetic Sequence analysis and
  Metabolic reconstructions of public genomes with
  Grid-based computational backend
• GNARE is PUMA2 for analysis of user-submitted
  genomes
• http//compbio.mcs.anl.gov/gnare)

• PUMA2 contains
• Integrates Information from over 25 genomic,
  metabolic, structural and taxonomic databases
  (RefSeq, Unirot, IproClass, PDB, KEGG, EMP, CATH,
  NCBI Taxonomy, Phenotypes, etc)
• Pre-computed analysis of publicly available
  completely and almost completely sequenced
  genomes (517 bacteria, 41 archaeal, 24
  eukaryotic, 638 mitochondrial and 2127 viral
  genomes) in interactive PUMA2 framework
• Automated Metabolic reconstructions for 300
  completely sequenced organisms
• GNARE User Models a framework for analysis of
  genomes provided by users (Shewanella federation,
  Apicomplexa genomes, strains of B. anthracis,
  Yersinia, Staphylococcus, Haemophilus, etc)
• A suite of unique tools for evolutionary analysis
  of enzymes and metabolic networks (Chisel,
  PhyloBlocks, etc) developed by our group
• PUMA2 satellite databases Pathos (GLRCE
  biodefence), TarGet (MCSG structural bilogy),
  Sentra (prokaryotic signal transduction),
  SubUnit, Physiological Profiles. MetaGenomes
  (PNNL Hanford Site), etc

6
GFF3 genomes repository at Argonneftp//ftp.mcs.a
nl.gov/pub/compbio/PUMA2/gff/gff_files/

• All completely sequenced genomes from RefSeq are
  converted into GFF3 format.
• GFF3 files for 8419 bacterial, eukaryotic,
  mitochondrial, viral, etc genomes can be
  downloaded from
• ftp//ftp.mcs.anl.gov/pub/compbio/PUMA2/gff/gff_f
  iles/
• The file names correspond to the NCBI-RefSeq
  accession numbers, e.g
• ftp//ftp.mcs.anl.gov/pub/compbio/PUMA2/gff/gff_fi
  les/NC_006815.gff

7
Future Plans Annotations

• In 2007 we will supplement Genome GFF3
  annotations for public genomes with
• additional annotations from public databases
  (e.g. NCBI, UniProt, Integr8, GenomeNet, etc) and
  
• annotations from our analysis tools (e.g. Chisel
  and PUMA2_FP), and other analysis tools
• Supplement the GFF3 files for RefSeq genomes with
  annotations provided by users via the GNARE system

8
Future Plans Sharing Annotations and
Cross-referencing the data

• GFF3 format can be used to share annotations and
  cross-references by different annotation centers
• We plan to build services (Web-services and
  Web-interfaces) to allow users to
• Submit and share their annotations via the PUMA2
  GFF3 converter
• Extract public annotations from PUMA2 integrated
  database as well as user-submitted annotations in
  GFF3 format
• Support customization of the GFF3 format (e.g.
  include only the fields of interest to a user,
  provide information from particular resource)
• Cross-references to various databases (e.g.
  NCBI-RefSeq, PIR, SwissProt, UniProt, and others)
  will be included as feature data in the GFF3
• Explore the use of ontologies for extension of
  the GFF3 format (we need your advice!)

9
Future Plans Data Distribution..(GFF3 genomes
repository at Argonne)

• All the feature data collected and computed by
  the PUMA2 project for publicly available genomes
  will be distributed in the GFF3 format.
• We will distribute the data through
• Web-services
• Web Interface (http)
• FTP downloads

10
Acknowledgements

• Our Team and
• Globus Ian Foster, Mike Wilde, Nika Nefedova,
  Jens Voeckler Condor Zach Miller, Miron Livny
  OSG, TeraGrid
• MCS Rick Stevens, systems, Susan Coghlan, and a
  lot of others.