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

Active learning with support vector machines in the drug discovery process.

Manfred K Warmuth1, Jun Liao, Gunnar Rätsch

  • 1Computer Science Department, University of California, Santa Cruz, California 95064, USA. manfred@cse.ucsc.edu

Journal of Chemical Information and Computer Sciences
|March 26, 2003
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

Striping artifact removal in VisiumHD data through nuclear counts modeling.

Bioinformatics (Oxford, England)·2026
Same author

Representation learning for multi-modal spatially resolved transcriptomics data.

Bioinformatics (Oxford, England)·2026
Same author

Estimation of Physiological Metrics from Resting ECGs Using Deep Learning in the UK Biobank, Including submaximal exercise derived V̇O <sub>2</sub> max, Body Fat Percentage, and Grip Strength.

medRxiv : the preprint server for health sciences·2026
Same author

RMS: a ML-based system for ICU respiratory monitoring and resource planning.

NPJ digital medicine·2025
Same author

ImmunoPepper: extracting personalized peptides from complex splicing graphs.

Bioinformatics (Oxford, England)·2025
Same author

Efficient and accurate search in petabase-scale sequence repositories.

Nature·2025

This study optimizes compound screening for drug design using active learning. Maximum margin hyperplanes from Support Vector Machines efficiently identify promising drug candidates, reducing biochemical testing.

Area of Science:

  • Computational chemistry
  • Machine learning
  • Bioinformatics

Background:

  • Drug discovery requires efficient screening of large compound libraries.
  • Minimizing biochemical testing iterations is crucial for cost and time efficiency.
  • Active learning offers a framework for iterative data acquisition and model improvement.

Purpose of the Study:

  • To apply active learning strategies for optimizing compound selection in computer-aided drug design.
  • To identify drug candidates that bind to a target molecule with minimal biochemical testing.
  • To compare the efficacy of different active learning selection strategies.

Main Methods:

  • Utilized the active learning paradigm for iterative compound batch selection.
  • Employed Support Vector Machines (SVM) to generate maximum margin hyperplanes.

Related Experiment Videos

  • The maximum margin hyperplane separates active from inactive compounds, maximizing distance to labeled data.
  • Conducted comparative studies on pharmaceutical datasets (DuPont Pharmaceuticals).

    • Main Results:

    • Active learning strategies based on maximum margin hyperplanes significantly outperformed simpler methods.
    • The SVM-based approach demonstrated superior efficiency in identifying binding compounds.
    • Demonstrated the effectiveness of this method in reducing the number of required biochemical tests.

    Conclusions:

    • Maximum margin hyperplane strategies derived from Support Vector Machines are highly effective for active learning in drug discovery.
    • This approach offers a computationally efficient and experimentally validated method for accelerating the identification of potential drug candidates.
    • The findings suggest a significant improvement in the efficiency of computer-aided drug design processes.