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 Concept Videos

You might also read

Related Articles

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

Sort by
Same author

[Identification of Placenta hominis and its adulterants using COI barcode].

Zhongguo Zhong yao za zhi = Zhongguo zhongyao zazhi = China journal of Chinese materia medica·2014
Same author

Two new species of Austrophthiracarus (Acari: Oribatida: phthiracaridae) from New Zealand.

Zootaxa·2014
Same author

The genus Notophthiracarus of New Zealand (Acari: Oribatida: Phthiracaridae): three new species and a key to 24 described species.

Zootaxa·2014
Same author

MHC class II restricted innate-like double negative T cells contribute to optimal primary and secondary immunity to Leishmania major.

PLoS pathogens·2014
Same author

Hepatic perfusion parameters of contrast-enhanced ultrasonography correlate with the severity of chronic liver disease.

Ultrasound in medicine & biology·2014
Same author

Dietary accumulation of tetrabromobisphenol A and its effects on the scallop Chlamys farreri.

Comparative biochemistry and physiology. Toxicology & pharmacology : CBP·2014

Related Experiment Video

Updated: Jul 19, 2025

A Protocol for Computer-Based Protein Structure and Function Prediction
16:41

A Protocol for Computer-Based Protein Structure and Function Prediction

Published on: November 3, 2011

68.7K

Estimating protein complex model accuracy based on ultrafast shape recognition and deep learning in CASP15.

Jun Liu1, Dong Liu1, Guangxing He1

  • 1College of Information Engineering, Zhejiang University of Technology, Hangzhou, China.

Proteins
|August 9, 2023
PubMed
Summary

DeepUMQA3 and GraphGPSM are new deep learning methods for protein complex accuracy estimation. DeepUMQA3 ranked first in CASP15 interface accuracy, outperforming other methods in predicting local distance difference test (lDDT).

Keywords:
CASPdeep learninginterface residuesmodel quality assessmentprotein complex structure

More Related Videos

High-Resolution Complexome Profiling by Cryoslicing BN-MS Analysis
09:33

High-Resolution Complexome Profiling by Cryoslicing BN-MS Analysis

Published on: October 15, 2019

7.3K
Exploring Caspase Mutations and Post-Translational Modification by Molecular Modeling Approaches
05:56

Exploring Caspase Mutations and Post-Translational Modification by Molecular Modeling Approaches

Published on: October 13, 2022

1.4K

Related Experiment Videos

Last Updated: Jul 19, 2025

A Protocol for Computer-Based Protein Structure and Function Prediction
16:41

A Protocol for Computer-Based Protein Structure and Function Prediction

Published on: November 3, 2011

68.7K
High-Resolution Complexome Profiling by Cryoslicing BN-MS Analysis
09:33

High-Resolution Complexome Profiling by Cryoslicing BN-MS Analysis

Published on: October 15, 2019

7.3K
Exploring Caspase Mutations and Post-Translational Modification by Molecular Modeling Approaches
05:56

Exploring Caspase Mutations and Post-Translational Modification by Molecular Modeling Approaches

Published on: October 13, 2022

1.4K

Area of Science:

  • Computational biology
  • Structural bioinformatics
  • Artificial intelligence in structural biology

Background:

  • Accurate prediction of protein complex structure is crucial for understanding biological functions.
  • Assessing the accuracy of predicted protein models is a key challenge in structural biology.
  • The Critical Assessment of techniques for protein Structure Prediction (CASP) is a benchmark for evaluating structure prediction methods.

Purpose of the Study:

  • To report and analyze the performance of novel deep learning methods, DeepUMQA3 and GraphGPSM, for protein complex model accuracy estimation.
  • To evaluate these methods in the 15th Critical Assessment of techniques for protein Structure Prediction (CASP15).
  • To introduce new deep learning approaches for characterizing multimeric protein complexes.

Main Methods:

  • Development of DeepUMQA3 and GraphGPSM, deep learning-based methods utilizing ensemble features from three levels: overall complex, intra-monomer, and inter-monomer.
  • Incorporation of ultrafast shape recognition (USR) for characterizing residue-topology relationships at overall and inter-monomer levels.
  • Application of deep residual and graph neural networks for accuracy estimation.

Main Results:

  • DeepUMQA3 achieved first place in interface residue accuracy estimation at CASP15, with a Pearson correlation of 0.570 for Local Distance Difference Test (lDDT), exceeding 0.5.
  • DeepUMQA3 demonstrated superior performance among top methods, achieving the highest Pearson correlation of lDDT on 25 out of 39 targets.
  • GraphGPSM showed strong performance in estimating overall fold accuracy, with TM-score Pearson correlations greater than 0.9 on 14 targets.

Conclusions:

  • DeepUMQA3 and GraphGPSM represent significant advancements in deep learning-based protein complex model accuracy estimation.
  • DeepUMQA3 is a highly accurate method for predicting interface residue accuracy, crucial for understanding protein-protein interactions.
  • The developed methods provide valuable tools for structural biologists, with publicly available servers for DeepUMQA3 and GraphGPSM.