Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Parallel BLAST on split databases.

David R Mathog1

  • 1Division of Biology, California Institute of Technology, Pasadena, CA 91125, USA. mathog@caltech.edu

Bioinformatics (Oxford, England)
|September 27, 2003
PubMed
Summary

This study presents a method for running BLAST programs on Beowulf clusters. Databases are split and distributed across nodes, enabling efficient parallel processing and comparable performance to large SMP machines.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Suppression of abdominal legs inDrosophila melanogaster.

Roux's archives of developmental biology : the official organ of the EDBO·2017
Same author

Evolution of the homeobox complex in the Diptera.

Current biology : CB·2003
See all related articles

Area of Science:

  • Bioinformatics
  • Computational Biology
  • High-Performance Computing

Background:

  • Large-scale sequence alignment often relies on Symmetric Multiprocessing (SMP) machines with ample memory for database caching.
  • Beowulf clusters offer a cost-effective alternative but typically lack sufficient memory per node to cache entire databases.

Purpose of the Study:

  • To adapt BLAST (Basic Local Alignment Search Tool) programs for efficient execution on Beowulf cluster configurations.
  • To achieve performance comparable to traditional SMP systems using distributed computing resources.

Main Methods:

  • Databases are partitioned into equally sized segments.
  • Each segment is stored locally on individual nodes within the Beowulf cluster.
  • BLAST queries are executed in parallel across all nodes.
  • Output files from each node are merged to produce a comprehensive final result.

Main Results:

  • The described program group enables Beowulf clusters to achieve performance comparable to large SMP machines for BLAST analysis.
  • This approach overcomes the memory limitations of individual nodes by leveraging the aggregate memory of the cluster.

Conclusions:

  • A novel strategy for distributed BLAST execution on Beowulf clusters has been developed.
  • This method facilitates high-performance sequence alignment on more accessible and scalable computing infrastructures.

Related Experiment Videos