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

Predicting mostly disordered proteins by using structure-unknown protein data.

Kana Shimizu1, Yoichi Muraoka, Shuichi Hirose

  • 1Department of Computer Science, Graduate School of Science and Engineering, Waseda University, 3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan. shimizu-kana@aist.go.jp

BMC Bioinformatics
|March 7, 2007
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

Knowledge and Attitudes about Dementia Patients: A Comparative Study between Japanese and Indonesian Dental Students.

Journal of International Society of Preventive & Community Dentistry·2026
Same author

c-FLIP: A pseudoprotease with emerging metal-binding activity.

The FEBS journal·2026
Same author

Accurate SPARQL generation via in-context learning and schema-based query construction.

Bioinformatics (Oxford, England)·2026
Same author

Nobiletin Prevents Cardiac Hypertrophy via SIRT5-Mediated Downregulation of p300.

Hypertension (Dallas, Tex. : 1979)·2025
Same author

Exploring Binding Sites on Proteins for Function Prediction Using the PoSSuM Databases.

Methods in molecular biology (Clifton, N.J.)·2025
Same author

Oligomer-based functions of mitochondrial porin.

Nature communications·2025

This study introduces a novel method for predicting intrinsically disordered proteins, outperforming existing approaches. The new technique effectively uses structure-unknown protein data, improving accuracy and robustness against sparse training datasets.

Area of Science:

  • Structural Biology
  • Bioinformatics

Background:

  • Intrinsically disordered proteins (IDPs) play crucial roles in cellular functions despite lacking stable structures.
  • A significant gap exists between known protein sequences and experimentally determined structures, with many sequences likely belonging to IDPs.
  • Leveraging structure-unknown protein data is essential to overcome training data sparseness in prediction models.

Purpose of the Study:

  • To develop a novel computational method for accurately predicting intrinsically disordered proteins.
  • To enhance prediction accuracy by incorporating information from structure-unknown protein sequences.
  • To improve the robustness of prediction models against variations in training data.

Main Methods:

  • A novel prediction method utilizing spectral graph transducer (SGT) was developed.

Related Experiment Videos

  • The model was trained on a large dataset combining structure-known and structure-unknown protein sequences.
  • Performance was evaluated against existing methods like FoldIndex, hydrophobicity-based approaches, support vector machines (SVMs), and per-residue predictors.

    • Main Results:

    • The SGT-based method demonstrated high sensitivity (0.723) and specificity (0.977) on a dataset of 82 disordered and 526 ordered proteins.
    • It achieved higher Matthews correlation coefficients compared to FoldIndex, hydrophobicity plots, and SVMs.
    • The method showed improved robustness against training data sparseness, with a higher correlation coefficient than SVMs.
    • When compared to per-residue predictors, the SGT method exhibited superior sensitivity and specificity for both disordered and ordered proteins.

    Conclusions:

    • The proposed spectral graph transducer-based method accurately predicts intrinsically disordered proteins.
    • Utilizing structure-unknown protein data enhances prediction accuracy and model robustness.
    • This approach effectively addresses the challenge of training data sparseness in protein disorder prediction.