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

Assessing scoring functions for protein-ligand interactions.

Philippe Ferrara1, Holger Gohlke, Daniel J Price

  • 1Department of Molecular Biology (TPC6), The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA.

Journal of Medicinal Chemistry
|May 28, 2004
PubMed
Summary

This study assesses nine scoring functions for molecular docking, finding they effectively identify correct protein-ligand poses but struggle to predict binding affinities accurately. Improved protonation state modeling enhances affinity predictions with certain 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

Accelerated Combinatorial Drug Design for Human Immunodeficiency Virus Resistance through Seeded Multisite λ-Dynamics.

Journal of chemical theory and computation·2026
Same author

Correction to "Identifying Artifacts from Large Library Docking".

Journal of medicinal chemistry·2025
Same author

Practically Significant Method Comparison Protocols for Machine Learning in Small Molecule Drug Discovery.

Journal of chemical information and modeling·2025
Same author

Knowledge and Structure-Based Drug Design of 15-PGDH Inhibitors.

Journal of medicinal chemistry·2025
Same author

Simultaneous Construction of Free Energy Surfaces via Multisite λ Dynamics and Umbrella Sampling.

Journal of chemical theory and computation·2025
Same author

Autobiography of Charles L. Brooks III.

The journal of physical chemistry. B·2025

Area of Science:

  • Computational Chemistry
  • Structural Biology
  • Drug Discovery

Background:

  • Molecular docking is crucial for identifying potential drug candidates by predicting protein-ligand interactions.
  • Scoring functions are essential for ranking docking predictions, but their accuracy varies.

Purpose of the Study:

  • To evaluate the performance of nine common scoring functions in molecular docking.
  • To assess their ability to recognize near-native protein-ligand conformations and predict binding affinities.
  • To investigate the impact of protonation states and solvation models on docking accuracy.

Main Methods:

  • Assessed nine scoring functions (CHARMm, DrugScore, AutoDock, DOCK, PMF, Gold, ChemScore) on 189 protein-ligand complexes.
  • Generated decoys using molecular dynamics and protein surface simulations.

Related Experiment Videos

  • Analyzed the influence of protonation states, solvation models (dielectric function, Generalized Born, Poisson equation), and rigid receptors.

    • Main Results:

    • Scoring functions showed good performance (around 80% recognition rate) in discriminating near-native from misdocked conformations.
    • Performance varied with decoy types (decoy vs. cross-decoy), with some functions degrading more significantly.
    • Predicting binding affinities remained challenging for all functions; ChemScore showed the highest correlation (R²=0.51).

    Conclusions:

    • The evaluated scoring functions are effective for pose recognition but less reliable for binding affinity prediction.
    • Protonation state modeling significantly impacts affinity prediction accuracy, especially with Generalized Born and Poisson models.
    • Further refinement of scoring functions is needed for accurate binding affinity prediction in drug discovery.