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

Quantitative structure-pharmacokinetic relationships for drug clearance by using statistical learning methods.

C W Yap1, Z R Li, Y Z Chen

  • 1Department of Computational Science, National University of Singapore, Blk SOC1, Level 7, 3 Science Drive 2, Singapore 117543, Singapore.

Journal of Molecular Graphics & Modelling
|November 18, 2005
PubMed
Summary

Quantitative structure-pharmacokinetic relationships (QSPkR) models predict drug clearance (CL(tot)). This study applies machine learning to 503 diverse compounds, achieving prediction accuracy comparable to earlier, less diverse models.

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

[Variables associated with severe cytopenia in adult chronic phase chronic myeloid leukemia patients receiving initial tyrosine kinase inhibitors].

Zhonghua xue ye xue za zhi = Zhonghua xueyexue zazhi·2026
Same author

Average and Heterogeneous Associations between Secondhand Smoking and Tooth Loss.

Journal of dental research·2026
Same author

[Proteomics analysis of the effect and mechanism of ADSCs on full-thickness skin defects in diabetic rats].

Zhonghua shao shang yu chuang mian xiu fu za zhi·2026
Same author

[Incidence and associated factors of tyrosine kinase inhibitor withdrawal syndrome and psychological issues in patients with chronic-phase chronic myeloid leukemia after therapy discontinuation].

Zhonghua xue ye xue za zhi = Zhonghua xueyexue zazhi·2025
Same author

[Plasmid characteristics and genome tracing analysis of a bacterial dysentery outbreak in Shandong Province, originating from <i>Shigella sonnei</i> producing extended spectrum β-lactamase].

Zhonghua yu fang yi xue za zhi [Chinese journal of preventive medicine]·2025
Same author

Physiological and molecular characteristics associated with the anti-senescence in <i>Camellia oleifera</i> Abel.

Photosynthetica·2024

Area of Science:

  • * Computational chemistry
  • * Cheminformatics
  • * Pharmacokinetics

Background:

  • * Quantitative structure-pharmacokinetic relationships (QSPkR) are vital for predicting drug lead pharmacokinetics.
  • * Existing QSPkR models for total clearance (CL(tot)) often use limited compound sets.
  • * A need exists for models applicable to a broader, more diverse range of compounds.

Purpose of the Study:

  • * To evaluate the performance of General Regression Neural Network (GRNN), Support Vector Regression (SVR), and K-Nearest Neighbour (KNN) for predicting CL(tot).
  • * To assess six distinct molecular descriptor sets for their predictive power in QSPkR models.
  • * To develop and validate QSPkR models using a large and diverse dataset of 503 compounds.

Main Methods:

  • * Applied GRNN, SVR, and KNN algorithms to model CL(tot) for 503 compounds.

Related Experiment Videos

  • * Utilized six molecular descriptor sets: DS-MIXED, DS-3DMoRSE, DS-ATS, DS-GETAWAY, DS-RDF, and DS-WHIM.
  • * Evaluated model performance using average-fold errors and the percentage of compounds within a two-fold prediction error.

    • Main Results:

    • * GRNN, SVR, and KNN models achieved average-fold errors between 1.63-2.23.
    • * 61.9-74.3% of compounds were predicted within a two-fold error of actual CL(tot) values.
    • * The DS-MIXED descriptor set, encompassing various chemical properties, generally yielded superior prediction accuracy.

    Conclusions:

    • * GRNN and SVR models demonstrate significant potential for predicting CL(tot) in diverse drug leads.
    • * The findings suggest that these machine learning methods can effectively extend QSPkR applications.
    • * Consensus models incorporating GRNN and SVR may offer enhanced predictive capabilities for pharmacokinetic properties.