Improved protein structure prediction using predicted interresidue orientations

2020/01/02 by Jianyi Yang, Ivan Anishchenko, Hahnbeom Park +3 · 27 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · Materials Science · #Computational Drug Discovery Methods #Enzyme Structure and Function #Protein Structure and Dynamics

paper · pdf · doi:10.1073/pnas.1914677117

openalex publication_date 2020/01/02 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/23

Abstract

The prediction of interresidue contacts and distances from coevolutionary data using deep learning has considerably advanced protein structure prediction. Here, we build on these advances by developing a deep residual network for predicting interresidue orientations, in addition to distances, and a Rosetta-constrained energy-minimization protocol for rapidly and accurately generating structure models guided by these restraints. In benchmark tests on 13th Community-Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction (CASP13)- and Continuous Automated Model Evaluation (CAMEO)-derived sets, the method outperforms all previously described structure-prediction methods. Although trained entirely on native proteins, the network consistently assigns higher probability to de novo-designed proteins, identifying the key fold-determining residues and providing an independent quantitative measure of the "ideality" of a protein structure. The method promises to be useful for a broad range of protein structure prediction and design problems.

Citations

• Continuous Automated Model EvaluatiOn (CAMEO) complementing the critical assessment of structure prediction in CASP12
• A further leap of improvement in tertiary structure prediction in CASP13 prompts new routes for future assessments
• Deep‐learning contact‐map guided protein structure prediction in CASP13

Cited by

• Recent advances in predicting and modeling protein–protein interactions
• CherryML: scalable maximum likelihood estimation of phylogenetic models
• Improving Signal and Transit Peptide Predictions Using AlphaFold2-predicted Protein Structures
• Construction of a Deep Neural Network Energy Function for Protein Physics
• Learning inverse folding from millions of predicted structures
• Evolutionary-scale prediction of atomic-level protein structure with a language model
• Physics‐based protein structure refinement in the era of artificial intelligence
• Highly accurate protein structure prediction for the human proteome
• A novel superfamily of bridge-like lipid transfer proteins
• Opportunities for artificial intelligence and synthetic biology in designing living drug delivery systems
• Triangle Multiplication Is All You Need For Biomolecular Structure Representations
• Atomically accurate de novo design of antibodies with RFdiffusion
• On the Potential of Machine Learning to Examine the Relationship Between Sequence, Structure, Dynamics and Function of Intrinsically Disordered Proteins
• ProteinBERT: a universal deep-learning model of protein sequence and function
• TBM preferred to AlphaFold 3 for functional models of insect odorant receptors
• Training Compute-Optimal Protein Language Models
• Reprogramming pretrained language models for protein sequence representation learning
• Computational design of serine hydrolases
• Recent Advances and Challenges in Protein Structure Prediction
• Exploring the Alternative Conformation of a Known Protein Structure Based on Contact Map Prediction
• Accurate prediction of protein structures and interactions using a three-track neural network
• Highly accurate protein structure prediction with AlphaFold
• ProFusion: 3D Reconstruction of Protein Complex Structures from Multi-view AFM Images
• Harnessing the power of artificial intelligence to advance cell therapy
• A guide to machine learning for biologists
• Madness, Cannibalism, and Traditional Fiction between Lu Xun and Mo Yan
• A structural biology community assessment of AlphaFold2 applications
• Pre-training Co-evolutionary Protein Representation via A Pairwise Masked Language Model

Related

• Advanced Deep Learning Methods for Protein Structure Prediction and Design
• Critical assessment of methods of protein structure prediction (CASP)—Round <scp>XIV</scp>
• Deep Learning-Driven Protein Structure Prediction and Design: Key Model Developments by Nobel Laureates and Multi-Domain Applications
• Accurate prediction of protein structures and interactions using a three-track neural network
• Highly accurate protein structure prediction with AlphaFold