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

Support vector machines in HTS data mining: Type I MetAPs inhibition study.

Jianwen Fang1, Yinghua Dong, Gerald H Lushington

  • 1Bioinformatics Core Facility, University of Kansas, Lawrence, KS 66045, USA. jwfang@ku.edu

Journal of Biomolecular Screening
|January 19, 2006
PubMed
Summary

Support vector machines (SVMs) effectively mine high-throughput screening data for drug discovery. SVM models identified 50% of active compounds by screening only 7% of the test set, improving hit rates.

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

Machine learning-based prognosis and early death prediction in de novo stage IV breast cancer patients with bone metastasis: a SEER database and multicentre retrospective study.

Scientific reports·2026
Same author

Syntheses of YCT-529: Net [3 + 2] Cycloaddition and One-Pot Bromination-Suzuki-Miyaura Coupling Strategies.

The Journal of organic chemistry·2026
Same author

Mechanisms for Anti-Inflammatory Activity of Gold Nanoparticles.

Combinatorial chemistry & high throughput screening·2026
Same author

Pharmacokinetic and pharmacological characterization of novel agmatine-based antihyperalgesic compounds.

The Journal of pharmacology and experimental therapeutics·2026
Same author

Sparse Pd-Te Covalent Bridges Drive Anomalous Bulk-to-Monolayer Electronic and Magnetic Evolution in FePd<sub>2</sub>Te<sub>2</sub>.

Advanced science (Weinheim, Baden-Wurttemberg, Germany)·2026
Same author

Development of TSSK1 and TSSK2 Targeted Degraders Reveals Sperm Machinery for Protein Degradation and Potential for Nonhormonal Male Contraception.

Journal of medicinal chemistry·2026

Area of Science:

  • Computational chemistry
  • Cheminformatics
  • Drug discovery

Background:

  • High-throughput screening (HTS) generates large datasets for identifying potential drug candidates.
  • Mining HTS data efficiently is crucial for accelerating drug discovery.
  • Type I methionine aminopeptidases (MetAPs) are targets in various therapeutic areas.

Purpose of the Study:

  • To apply Support Vector Machines (SVMs) for analyzing HTS data in a MetAP inhibition study.
  • To evaluate the performance of SVM models in prioritizing active compounds from a large chemical library.
  • To introduce and utilize the PT50 score for assessing model efficiency in hit identification.

Main Methods:

  • A library of 43,736 small organic molecules was screened for MetAP inhibition.

Related Experiment Videos

  • Support Vector Machines (SVMs) were employed to build predictive models from a training set.
  • Compounds were ranked based on SVM decision values, and a novel PT50 score was calculated.
  • The dataset was randomly split into training and test sets (3:1 ratio).

    • Main Results:

    • SVM models successfully ranked compounds in the test set, significantly increasing hit rates.
    • 50% of active compounds were recovered by screening only 7% of the test set using optimized SVM models.
    • The size of the training set was found to be critical for model performance, with 10,000 compounds suggested as a minimum.

    Conclusions:

    • SVMs provide a powerful computational approach for efficient analysis of HTS data.
    • The PT50 metric effectively measures the performance of predictive models in hit identification.
    • Careful selection of model parameters and sufficient training data are essential for successful application of SVMs in drug discovery.