Toward the solution of the protein structure prediction problem

Pearce, Robin; Zhang, Yang

doi:10.1016/j.jbc.2021.100870

Cited by 97 publications

(123 citation statements)

References 193 publications

Supporting

Mentioning

120

Contrasting

Unclassified

Order By: Relevance

“…TBM, also termed homology modeling or comparative modeling, usually consists of the following steps 32 , 33 : 1- adopting the homologous proteins with experimentally defined structures as templates; 2- alignment of the target sequences and templates; 3- building the 3D model, based on the alignment; and 4- refinement of the models. Current methods regularly treat each step distinctly, and the full TBM process can then be set up by merging methods for each of the above steps 31 .…”

Section: Methodsmentioning

confidence: 99%

Designing a Multi-Epitope Antigen for Serodiagnosis of Strongyloides stercoralis Based on L3Nie.01 and IgG Immunoreactive Epitopes

Movahedpour¹,

Mostafavi‐Pour²,

Sarkari³

et al. 2022

AJMB

View full text Add to dashboard Cite

Background: Serological diagnosis of Strongyloides stercoralis (S. stercoralis) is frequently challenging because of cross-reactivity with other parasitic nematodes. Therefore, it is necessary to introduce novel serological tests with high performance to properly diagnose this neglected parasitic infection. The purpose of the current study was to design a multi-epitope construct for the diagnosis of S. stercoralis. Methods: For the purpose of this study, first, highly antigenic segments and potential immunodominant epitopes of S. stercoralis were identified from two antigenic proteins, and then all of the selected parts were linked by an appropriate linker. Next, the physico-chemical features of the designed construct were analyzed. Then, tertiary structures of the construct were built and evaluated to find out the best one. Lastly, the amino acid sequence was reverse-translated and optimized for over-expression in Escherchia coli (E. coli). Results: The bioinformatic evaluation indicated that the designed protein construct could be hydrophilic, thermostable, and acidic and the estimated half-life was more than 10 hr in E. coli. Conclusion: According to the results of the study, the designed construct could be used as an efficient antigen in the ELISA system for serological diagnosis of human strongyloidiasis.

show abstract

Section: Methodsmentioning

confidence: 99%

Designing a Multi-Epitope Antigen for Serodiagnosis of Strongyloides stercoralis Based on L3Nie.01 and IgG Immunoreactive Epitopes

Movahedpour¹,

Mostafavi‐Pour²,

Sarkari³

et al. 2022

AJMB

View full text Add to dashboard Cite

show abstract

“…To improve the performance of RNA structure prediction methods, we drew inspiration from recent advances in protein structure prediction, where deep learning techniques have revolutionized the field [15][16][17] . Toward this goal, we developed DeepFoldRNA, which uses a self-attention-based neural network architecture to predict geometric restraints, where 3D RNA structures are then built using limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) minimization simulations.…”

Section: Introductionmentioning

confidence: 99%

“…Recently, deep learning approaches have been successfully applied to the problem of model selection 14 . Nevertheless, the success of these methods is predicated on generating conformations that are close to the native structures, where atomic resolution was only obtained after utilizing restraints from native structures, which are not available in practical modeling applications.To improve the performance of RNA structure prediction methods, we drew inspiration from recent advances in protein structure prediction, where deep learning techniques have revolutionized the field [15][16][17] . Toward this goal, we developed DeepFoldRNA, which uses a self-attention-based neural network architecture to predict geometric restraints, where 3D RNA structures are then built using limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) minimization simulations.…”

mentioning

confidence: 99%

De NovoRNA Tertiary Structure Prediction at Atomic Resolution Using Geometric Potentials from Deep Learning

Pearce

Omenn

Zhang

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Experimental characterization of RNA structure remains difficult, especially for non-coding RNAs that are critical to many cellular activities. We developed DeepFoldRNA to predict RNA structures from sequence alone by coupling deep self-attention neural networks with gradient-based folding simulations. The method was tested on two independent benchmark datasets from Rfam families and RNA-Puzzle experiments, where DeepFoldRNA constructed models with an average RMSD=2.69 Å and TM-score=0.743, which outperformed state-of-the-art methods and the best models submitted from the RNA-Puzzles community by a large margin. On average, DeepFoldRNA required ~1 minute to fold medium-sized RNAs, which was ~350-4000 times faster than the leading Monte Carlo simulation approaches. These results demonstrate the major advantage of advanced deep learning techniques to learn more accurate information from evolutionary profiles than knowledge-based potentials derived from simple statistics of the PDB library. The high speed and accuracy of the developed method should enable large-scale atomic-level RNA structure modeling applications.

show abstract

“…Protein folding has been a major focus in biochemistry and computational research in recent decades, motivated by its central role in protein function and protein homeostasis [20,45]. From a computational perspective, protein structure prediction methods can be broadly classified into knowledge-based and physics-based approaches, along with more recent deep learning algorithms [46]. Indeed, the field has experienced a revolution with the publication of AlphaFold2, an AI algorithm capable of predicting the 3D apo structure of single protein domains and multimeric systems with experimental accuracy [35,25].…”

Section: Introductionmentioning

confidence: 99%

Peptide conformational sampling using the Quantum Approximate Optimization Algorithm

Boulebnane¹,

Lucas²,

Meyder³

et al. 2022

Preprint

View full text Add to dashboard Cite

Protein folding -the problem of predicting the spatial structure of a protein given its sequence of amino-acids -has attracted considerable research effort in biochemistry in recent decades. In this work, we explore the potential of quantum computing to solve a simplified version of protein folding. More precisely, we numerically investigate the performance of a variational quantum algorithm, the Quantum Approximate Optimization Algorithm (QAOA), in sampling low-energy conformations of short peptides. We start by benchmarking the algorithm on an even simpler problem: sampling self-avoiding walks, which is a necessary condition for a valid protein conformation. Motivated by promising results achieved by QAOA on this problem, we then apply the algorithm to a more complete version of protein folding, including a simplified physical potential. In this case, based on numerical simulations on 20 qubits, we find less promising results: deep quantum circuits are required to achieve accurate results, and the performance of QAOA can be matched by random sampling up to a small overhead. Overall, these results cast serious doubt on the ability of QAOA to address the protein folding problem in the near term, even in an extremely simplified setting. We believe that the approach and conclusions presented in this work could offer valuable methodological insights on how to systematically evaluate variational quantum optimization algorithms on real-world problems beyond protein folding.

show abstract

Toward the solution of the protein structure prediction problem

Cited by 97 publications

References 193 publications

Designing a Multi-Epitope Antigen for Serodiagnosis of Strongyloides stercoralis Based on L3Nie.01 and IgG Immunoreactive Epitopes

Designing a Multi-Epitope Antigen for Serodiagnosis of Strongyloides stercoralis Based on L3Nie.01 and IgG Immunoreactive Epitopes

De NovoRNA Tertiary Structure Prediction at Atomic Resolution Using Geometric Potentials from Deep Learning

Peptide conformational sampling using the Quantum Approximate Optimization Algorithm

Contact Info

Product

Resources

About