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 Video

Updated: Jan 23, 2026

Observing the Transformation of Bodily Self-consciousness in the Squeeze-machine Experiment
07:20

Observing the Transformation of Bodily Self-consciousness in the Squeeze-machine Experiment

Published on: March 8, 2019

14.2K

Value of p-Value.

Alexander Golbraikh1

  • 1Laboratory for Molecular Modeling, University of North Carolina at Chapel Hill, CB #7360, Chapel Hill, NC 27599.

Molecular Informatics
|June 13, 2019
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 journal

SpaceExpander: An Automated System for Drafting Markush Claims to Expand Chemical Space.

Molecular informatics·2026
Same journal

A Structure-Informed Atlas of Venom-Derived Peptides Reveals the Organization of Chemical Space.

Molecular informatics·2026
Same journal

ConGen: Targeted Molecule Generation Through Contrastive Learning and Latent Optimization.

Molecular informatics·2026
Same journal

Novel Molecules Generation Using Graph Generative Adversarial Networks.

Molecular informatics·2026
Same journal

An Attention-Driven Graph Transformer With Nonlinear Modeling and Neuro-Fuzzy Fusion for High-Order Toxic Molecular Graph Learning.

Molecular informatics·2026
Same journal

Molecular Modeling and Chemoinformatics in Ukraine.

Molecular informatics·2026
See all related articles

External validation of Quantitative Structure-Activity/Property/Toxicity Relationship (QS/A/P/T/R) models requires careful consideration of dataset size. This study proposes thresholds for using p-value or classification accuracy criteria for reliable model assessment.

Area of Science:

  • * Cheminformatics
  • * Computational Chemistry
  • * Data Science

Background:

  • * External validation is crucial for assessing the reliability of Quantitative Structure-Activity/Property/Toxicity Relationship (QS/A/P/T/R) models.
  • * Existing literature often overlooks specific criteria for validating classification and category-based QS/A/P/T/R models.
  • * Key metrics for external validation include statistical significance (p-value) and predictive accuracy (Correct Classification Rate - CCR).

Purpose of the Study:

  • * To address critical, under-discussed aspects of external validation for classification and category QS/A/P/T/R models.
  • * To establish guidelines for selecting appropriate validation criteria based on the number of compounds in external validation sets.
  • * To ensure models are both statistically significant and predictive for each class or category.
Keywords:
classification and category response variablesexternal validation of QSAR modelsp-values and correct classification rate as alternative criteria of prediction accuracy

More Related Videos

Spinal Cord Transection In Xenopus laevis Tadpoles
05:54

Spinal Cord Transection In Xenopus laevis Tadpoles

Published on: December 10, 2021

4.5K
Myocardial Infarction and Functional Outcome Assessment in Pigs
12:03

Myocardial Infarction and Functional Outcome Assessment in Pigs

Published on: April 25, 2014

28.6K

Related Experiment Videos

Last Updated: Jan 23, 2026

Observing the Transformation of Bodily Self-consciousness in the Squeeze-machine Experiment
07:20

Observing the Transformation of Bodily Self-consciousness in the Squeeze-machine Experiment

Published on: March 8, 2019

14.2K
Spinal Cord Transection In Xenopus laevis Tadpoles
05:54

Spinal Cord Transection In Xenopus laevis Tadpoles

Published on: December 10, 2021

4.5K
Myocardial Infarction and Functional Outcome Assessment in Pigs
12:03

Myocardial Infarction and Functional Outcome Assessment in Pigs

Published on: April 25, 2014

28.6K

Main Methods:

  • * Analysis of the relationship between dataset size, p-value, and Correct Classification Rate (CCR).
  • * Definition of three distinct thresholds for the number of compounds per class in external validation sets.
  • * Proposed decision framework for applying p-value or CCR criteria based on these thresholds.

Main Results:

  • * The p-value criterion is unattainable below a specific compound number threshold.
  • * For intermediate dataset sizes, the choice between p-value and CCR depends on the p-value itself.
  • * As dataset size increases, CCR becomes the primary criterion, with prediction error converging to expected error.

Conclusions:

  • * The number of compounds in external validation sets dictates the appropriate assessment metric (p-value vs. CCR).
  • * Implementing these thresholds ensures more robust and reliable validation of classification/category QS/A/P/T/R models.
  • * Findings have implications for multidimensional data analysis beyond cheminformatics.