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

Nonparametric regression applied to quantitative structure-activity relationships

Constans1, Hirst

  • 1Department of Molecular Biology, Scripps Research Institute, La Jolla, California 92037, USA.

Journal of Chemical Information and Computer Sciences
|April 13, 2000
PubMed
Summary

Additive nonparametric regressors demonstrated superior predictive accuracy for quantitative structure-activity relationship (QSAR) modeling compared to multilinear regression (MLR). Principal components did not enhance performance with the current descriptors.

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

Army Medical Department.

Edinburgh medical journal·2018
Same author

Army Medical Department.

Edinburgh medical journal·2018
Same author

Computing power revolution and new algorithms: GP-GPUs, clouds and more: general discussion.

Faraday discussions·2014
Same author

Applications and serious games: from docking to protein folding: general discussion.

Faraday discussions·2014
Same author

Advanced visualization and visual analytics: general discussion.

Faraday discussions·2014
Same author

Virtual and augmented reality immersive molecular simulations: general discussion.

Faraday discussions·2014

Area of Science:

  • Medicinal Chemistry
  • Computational Chemistry
  • Cheminformatics

Background:

  • Quantitative Structure-Activity Relationship (QSAR) modeling is crucial for drug discovery.
  • Nonparametric regression methods offer flexible alternatives to traditional linear models.
  • Evaluating advanced regressors is essential for improving predictive accuracy in QSAR.

Purpose of the Study:

  • To assess the performance of various nonparametric regressors for QSAR modeling.
  • To compare nonparametric methods against linear (MLR) and nonlinear (smoothing splines) benchmarks.
  • To investigate the impact of variable selection and principal components on QSAR model performance.

Main Methods:

  • Nadaraya-Watson, local linear, and shifted Nadaraya-Watson regressors were implemented.

Related Experiment Videos

  • Additive models were used for computational efficiency with local linear and shifted Nadaraya-Watson.
  • Performance was benchmarked using mean absolute error and cross-validation correlation against MLR and smoothing splines.

    • Main Results:

    • Additive nonparametric regressors exhibited greater predictive accuracy than MLR for dopamine beta-hydroxylase inhibitors.
    • Principal components did not improve nonparametric regressor performance due to low correlation among original descriptors.
    • The tested nonparametric methods outperformed MLR in predictive power.

    Conclusions:

    • Additive nonparametric regressors are effective for QSAR modeling, outperforming MLR.
    • Further research is needed to integrate nonparametric methods with advanced variable selection and dimensionality reduction for high-dimensional QSAR.
    • Nonparametric approaches show promise for enhancing QSAR predictive capabilities.