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

Validation tools for variable subset regression.

Knut Baumann1, Nikolaus Stiefl

  • 1Department of Pharmacy and Food Chemistry, University of Wuerzburg, Am Hubland, D-97074 Wuerzburg, Germany. knut.baumann@mail.uni-wuerzburg.de

Journal of Computer-Aided Molecular Design
|February 26, 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

AbDist: a lightweight, distance-based model for antibody affinity prediction as an interpretable benchmark for machine learning models.

mAbs·2026
Same author

Improvement of Peak Integration in Capillary Electrophoresis: Reference Data Set No. 1.

Electrophoresis·2026
Same author

Discovery and Characterization of Zilurgisertib, a Potent and Selective Inhibitor of Activin Receptor-like Kinase‑2 (ALK2) for the Treatment of Fibrodysplasia Ossificans Progressiva.

ACS medicinal chemistry letters·2025
Same author

The changing landscape of medicinal chemistry optimization.

Nature reviews. Drug discovery·2025
Same author

Discovery of NP3-253, a Potent Brain Penetrant Inhibitor of the NLRP3 Inflammasome.

Journal of medicinal chemistry·2024
Same author

Balancing Molecular Size, Activity, Permeability, and Other Properties: Drug Candidates in the Context of Their Chemical Structure Optimization.

Journal of chemical information and modeling·2024
Same journal

RANQSAR: a standalone open-source application for reproducible machine learning-based QSAR analysis.

Journal of computer-aided molecular design·2026
Same journal

Integrating evolutionary and compositional features with ML and DL for robust and interpretable druggable protein prediction.

Journal of computer-aided molecular design·2026
Same journal

QUAD: a composite risk framework integrating uncertainty, applicability domain, and model disagreement for reliable QSAR predictions.

Journal of computer-aided molecular design·2026
Same journal

Comparative quantum-chemical investigation of 2-chloro-N-(4-methoxyphenyl)acetamide and 2-(4-methoxyphenylamino)-2-oxoethyl meth/acrylate: DFT, TD-DFT, and non-covalent interaction analyses.

Journal of computer-aided molecular design·2026
Same journal

Discovery of a novel CaMKⅡα inhibitor by machine learning, molecular docking and molecular dynamics simulation.

Journal of computer-aided molecular design·2026
Same journal

Explainable QSAR models of 5-HT1A receptor ligands using conceptual DFT descriptors and no-code machine learning tools.

Journal of computer-aided molecular design·2026
See all related articles

A new validation protocol for quantitative structure-activity relationship (QSAR) models ensures reliable variable selection. This method enhances model stability and predictive accuracy, crucial for trustworthy QSAR research.

Area of Science:

  • Quantitative Structure-Activity Relationship (QSAR) studies
  • Cheminformatics
  • Computational Chemistry

Background:

  • Variable selection is a critical step in QSAR research, significantly impacting model performance and reliability.
  • Thorough validation of variable selection techniques is essential to ensure the trustworthiness of QSAR models.

Purpose of the Study:

  • To introduce a robust validation protocol for QSAR variable selection methods.
  • To present two novel tools for assessing model complexity, stability, and internal performance metrics.

Main Methods:

  • Development of a validation protocol incorporating permutation testing and noise addition techniques.
  • Assessment of internal figures of merit, such as cross-validated prediction error, using permutation testing.

Related Experiment Videos

  • Determination of model complexity and stability through noise addition, evaluating models generated by variable selection.

    • Main Results:

    • Permutation testing effectively assesses the inflation of internal QSAR model performance metrics.
    • Noise addition quantifies model complexity and stability, aiding in the evaluation of variable selection outcomes.
    • Leave-multiple-out cross-validation was confirmed to produce more stable QSAR models compared to leave-one-out.
    • The proposed validation protocol demonstrated good internal and external model performance across eight diverse QSAR datasets.

    Conclusions:

    • The presented validation protocol provides reliable statistical information for assessing QSAR model predictive abilities.
    • Graphical outputs from the validation tools offer an accessible and reliable assessment of model performance.
    • The protocol is effective in evaluating the impact of different cross-validation strategies on QSAR model characteristics.