GNARE-GADU: An Environment for Grid-Based High-Throughput Genome Analysis

1
GNARE-GADU An Environment for Grid-Based
High-Throughput Genome Analysis

• Mathematics and Computer Science Division
• Argonne National Laboratory
• Natalia Maltsev, Dina Sulakhe,
• Alex Rodriguez Bioinformatics group
• Mike Wilde, Nika Nefedova, Jens Voeckler
• Ian Foster the Globus Group

2
MCS Argonne Bioinformatics Group
Projectshttp//compbio.mcs.anl.gov
PUMA2 a Grid-enabled interactive environment
for analysis of evolution of metabolism
http//compbio.mcs.anl.gov/puma2 PUMA2 contains
analysis of 1032 genomes, Integrates
information from over 20 sequence and metabolic
databases Supports user annotations Over 3500
unique individual users in April 2005

• Application Projects depending on
  PUMA2/GADU/GNARE
• Pathos Microbial informatics Core for NIH
  Great Lakes Center of Excellence in Biodefense
  and emerging infections
• TarGet NIH Midwest Structural Biology Center
• MetaGenomes DOE Microbial Genomes program
• Sentra -- database prokaryotic signal
  transduction
• SubUnit database
• Physiological Profiles
• PUMA2 tools
• CHISEL a workbench for evolutionary analysis of
  enzymes,assignment of functions to genes,
  PhyloBlocks, BloBla, SVMMER, etc

• GADU/GNARE provides PUMA2 with
• Grid-based computational Backend
• Controlled automated workflows for data
  acquisition and analysis

3
Biggest Challenge !! Site Selection.
Gadu runs jobs dynamically on different grid
sites without reserving any nodes. Its difficult
to maintain the status of all the sites and
select a site that is appropriate for a given job.
4
GADU/Gnare Future Plans and Use

• Alpha testing with the individual projects.
• Ongoing Projects
• SEED
• PNNL-Microbial Genome Program
• Northern Illinois University Rice Genome
  Project.
• Northwestern University
• Web Services
• Use of Grid infrastructure
• Goal is of 2 databases updates per month
  (includes Blast, Blocks)
• Each database update entails 20000 jobs running
  at about 2 hours each on one node.