Efforts to Link Ecological Metadata with Bacterial Gene Sequences at the Sapelo Island Microbial Observatory

1
Efforts to Link Ecological Metadata with
Bacterial Gene Sequences at theSapelo Island
Microbial Observatory

• Wade M. Sheldon
• Mary Ann Moran
• James T. Hollibaugh

2
Genetic Sequence Databases

• Major informatics success story
• Large repositories for nucleotide sequences (e.g.
  GenBank/EMBL/NDDJ 16M)
• Automated and web-based data submission -
  required as part of publication process
• Standardized alignment/search tools support use
  for classification
• Numerous environmental sequences ecologists
  now using to study biogeography, community
  structure, eco-physiology

3
Problems with GenBank

• Metadata voluntary limited in scope
• Title (definition), authors, key words, comments,
  literature citation
• Many sequences unpublished, undescribed
• Quality control standards poorly enforced
• No direct way to provide links to ancillary data
  (URLs not officially supported, often removed)
• Very inefficient and often impossible for
  investigators to obtain ecological context
  information, even from journals
• Comparisons of matched taxa by traits not possible

4
Consequence

• Tremendous amount of bacterial sequence data
  relevant to microbial ecologists
• No established interface

5
Example Insufficient Metadata
6
Sapelo Island Microbial Observatory
(http//simo.marsci.uga.edu)

• MObs NSF-funded network of sites or "microbial
  observatories" established to discover novel
  microorganisms, microbial consortia, communities,
  activities and other novel properties, and to
  study their roles in diverse environments
• Projects supported are expected to establish or
  participate in an established, Internet-accessible
  knowledge network to disseminate the information
  resulting from these activities
• SIMO - Investigating the diversity of
  prokaryotes, their physiological and genetic
  characteristics, and their biogeochemical
  activities in a salt marsh/estuarine ecosystem in
  the southeastern U.S.
• Knowledge networks
• GenBank
• GCE-LTER IS
• SIMO 16S rRNA Database

7
SIMO 16S rRNA Database

• Purpose LIMS, research tool, data dissemination
• Designed to store sequence data and all
  supporting SIMO research information
• Hierarchical structure modeled after research
  workflow
• Metadata on site geography, sample collection,
  all methodology, personnel, ancillary
  measurements
• Extensive content control, error checking
• Links to information in external databases (RDP
  II, GenBank, GCE-LTER)
• Queries by phylogenic and/or ecological
  characteristics

8
Conceptual Diagram of the SIMO Database
9
List-based data entry linked to metadata tables
10
Controlled vocabulary supports finely-targeted
queriesAutomatic hyperlinks provide links to
tasks
11
List-based queries also simplify public interface
12
Phylogenetic and ecological characteristics
combined dynamically to create overview and query
interface
13
SIMO Metadata

• Metadata primarily stored in managed lists,
  linked to records by foreign key fields
• Scalable design details can be added
  independently without altering data records
• Complete metadata for sequences generated by
  relational joins
• Links to external metadata in GCE-LTER database
  adds site geography, research history, long-term
  environmental characteristics

14
Metadata Standards

• No existing standard for environmental sequence
  metadata
• Sequence formats (FASTA, BIOML, BSML) designed
  for data parsing, sequence annotation
• SIMO metadata currently displayed in summary form
  on sequence detail pages
• Exploring adopting emerging standards like EML

15
Sequence Details
16
Future Directions

• Incorporating batch upload features for library
  submissions
• Integrating database with RDP SeqMatch Agent
  programs for automatic phylogenetic analysis,
  sequence annotation
• Provide full metadata in formatted/printable and
  parsable ASCII formats (XML)
• Participate in Entrez Link-Out to provide links
  to SIMO sequence entries from GenBank