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

Could statistical potential models achieve comparable or better performance than deep learning models?

Zhihao Wang1, Sheng Wang2, Jingjing Guo3

  • 1School of Physics, Shandong University, 27 Shanda Nan Road, 250100 Jinan, Shandong Province, China.

Briefings in Bioinformatics
|March 2, 2026
PubMed
Summary

Related Concept Videos

Pharmacokinetic Models: Comparison and Selection Criterion01:26

Pharmacokinetic Models: Comparison and Selection Criterion

429
Physiological and compartmental models are valuable tools used in studying biological systems. These models rely on differential equations to maintain mass balance within the system, ensuring an accurate representation of the dynamic processes at play.
Physiological models take a detailed approach by considering specific molecular processes. They can predict drug distribution, metabolism, and elimination changes, providing a comprehensive understanding of how drugs interact with the body.
429

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Tonoplast sugar transporter StTST1 mediates vacuolar sugar partitioning and abiotic stress tolerance in potato.

Plant physiology and biochemistry : PPB·2025
Same author

Physical Properties of Mold Flux and Mineralogical Characteristics of Flux Film for Low-alloy Peritectic Steel Continuous Casting.

Materials (Basel, Switzerland)·2025
Same author

Regulation of osteogenic differentiation in vascular smooth muscle cells under high-glucose condition.

Frontiers in endocrinology·2025
Same author

Accurate structure prediction of cyclic peptides containing unnatural amino acids using HighFold3.

Briefings in bioinformatics·2025
Same author

Silibinin accelerates diabetic wound healing through PI3K/Akt-mediated immunomodulation-angiogenesis crosstalk.

Biochemical and biophysical research communications·2025
Same author

YTHDC2 inhibits HPV-positive cervical cancer growth by suppressing SLC7A11 in a ferroptosis-dependent manner.

Oncology letters·2025
Same journal

STED: flexible cross-modal topic modeling infers cell-type-specific regulatory landscapes from bulk epigenomics.

Briefings in bioinformatics·2026
Same journal

A knowledge-guided deep learning framework for quantitative nucleic acid testing.

Briefings in bioinformatics·2026
Same journal

Optimal transport for label transfer in single-cell multi-omics integration.

Briefings in bioinformatics·2026
Same journal

Continuous multi-omics pathway enrichment analysis resolves hidden functional heterogeneity.

Briefings in bioinformatics·2026
Same journal

Evaluating completeness, coherence, and consistency of genome-scale function annotations.

Briefings in bioinformatics·2026
Same journal

Transformers for single-cell RNA sequencing: a survey.

Briefings in bioinformatics·2026
See all related articles
This summary is machine-generated.

Statistical potentials offer an efficient alternative for protein-ligand interaction prediction in drug discovery. HybridSP, a novel statistical potential, rivals deep learning models in docking and virtual screening accuracy.

Area of Science:

  • Computational chemistry
  • Structural biology
  • Drug discovery

Background:

  • Accurate prediction of protein-ligand interactions is crucial for structure-based drug discovery.
  • Deep learning models show promise, but traditional statistical potentials are underexplored, especially with limited data.

Purpose of the Study:

  • To systematically assess statistical potential models for protein-ligand docking and virtual screening.
  • To develop a hybrid statistical potential model that combines the strengths of different potential types.

Main Methods:

  • Evaluation of various statistical potential models, including distance-dependent pairwise atom-atom and orientation-dependent atom-residue potentials.
  • Development and application of HybridSP, a novel hybrid potential integrating multiple terms.
Keywords:
protein–ligand interactionscoring functionstatistical potential

Related Experiment Videos

  • Affinity-weighted scheme for bias correction in statistical distributions.
  • Validation on CASF-2016, DUD-E, and DUD-A benchmarks.

    • Main Results:

    • Docking performance is enhanced by distance-dependent pairwise atom-atom potentials.
    • Virtual screening benefits from orientation-dependent atom-residue potentials.
    • HybridSP achieved a 91.6% docking success rate and a 29.35 enrichment factor at top 1% on CASF-2016.
    • HybridSP demonstrated strong virtual screening capabilities on DUD-E and DUD-A datasets.

    Conclusions:

    • Well-designed statistical potentials can achieve high performance and interpretability without complex deep learning architectures.
    • HybridSP offers an efficient and effective alternative for scoring function design in drug discovery.
    • The developed models provide a viable option for data-limited scenarios in computational chemistry.