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

A linear memory algorithm for Baum-Welch training.

István Miklós1, Irmtraud M Meyer

  • 1MTA-ELTE Theoretical Biology and Ecology Group, Pázmány Péter sétány 1/c 1117 Budapest, Hungary. miklosi@ramet.elte.hu

BMC Bioinformatics
|September 21, 2005
PubMed
Summary
This summary is machine-generated.

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

Splicing dependency between EIF4G2 introns is mediated by exon definition and relies on a downstream splicing event.

The Journal of biological chemistry·2026
Same author

DuplexDiscoverer: a computational method for the analysis of experimental duplex RNA-RNA interaction data.

Nucleic acids research·2025
Same author

Constructing and sampling partite, 3-uniform hypergraphs with given degree sequence.

PloS one·2024
Same author

e-RNA: a collection of web-servers for the prediction and visualisation of RNA secondary structure and their functional features.

Nucleic acids research·2023
Same author

CYCLeR-a novel tool for the full isoform assembly and quantification of circRNAs.

Nucleic acids research·2022
Same author

Bacteriophages: Emerging Applications in Medicine, Food, and Biotechnology.

PHAGE (New Rochelle, N.Y.)·2022
Same journal

OpenIMC: an open-source platform for analyzing single-cell and spatial proteomics by imaging mass cytometry.

BMC bioinformatics·2026
Same journal

NAP: an open source pipeline for cross-domain microbiome profiling using Nanopore sequencing-derived amplicon data.

BMC bioinformatics·2026
Same journal

SurvGME: an R package for survival analysis with graphical and measurement error models.

BMC bioinformatics·2026
Same journal

SimMapNet: a Bayesian framework for gene regulatory network inference using gene ontology similarities as external hint.

BMC bioinformatics·2026
Same journal

Dual channel drug-drug interactions extraction based on cross attention.

BMC bioinformatics·2026
Same journal

FeSseqdb: a curated sequence-level database and interpretable machine learning framework for identifying iron-sulfur proteins.

BMC bioinformatics·2026
See all related articles

We developed a new linear space algorithm for Baum-Welch training, significantly reducing memory usage for hidden Markov models. This breakthrough enables efficient training of complex models, advancing biological sequence analysis.

Area of Science:

  • Computational Biology
  • Algorithm Development
  • Bioinformatics

Background:

  • Baum-Welch training is an expectation-maximization algorithm for hidden Markov models (HMMs).
  • Current algorithms are memory-intensive, hindering HMM application in biological sequence analysis.
  • Existing methods struggle with long sequences and complex models like pair HMMs.

Purpose of the Study:

  • To introduce a novel, memory-efficient algorithm for Baum-Welch training.
  • To overcome the memory limitations of existing Baum-Welch training algorithms.
  • To enable automatic parameter training for complex HMMs in biological sequence analysis.

Main Methods:

  • Developed a linear space algorithm for Baum-Welch training.
  • Algorithm achieves O(M) memory and O(LMTmax(T+E)) time per iteration for single sequences.

Related Experiment Videos

  • Algorithm offers memory-time trade-offs for n-HMMs and n sequences.

    • Main Results:

    • The new algorithm reduces memory requirement independently of training sequence length.
    • Achieved linear space complexity, a significant improvement over previous O(log(L)M) methods.
    • For n-HMMs, memory reduced from O(log(L)L(n-1)M) to O(L(n-1)M).

    Conclusions:

    • The novel algorithm is both faster and more memory-efficient for HMMs used in gene prediction.
    • Enables efficient parameter training for large-scale biological sequence analyses.
    • Removes memory constraints previously limiting HMM applications.