Journal Home > Volume 27 , Issue 5
Tsinghua Science and Technology 2022, 27(5): 843-854 https://doi.org/10.26599/TST.2021.9010064
Open Access | Issue | Published: 17 March 2022

Protein Residue Contact Prediction Based on Deep Learning and Massive Statistical Features from Multi-Sequence Alignment

Show Author's Information Huiling Zhang†,  Min Hao†,  Hao Wu†,  Hing-Fung Ting,  Yihong Tang( ),  Wenhui Xi( ),  Yanjie Wei( )
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China
University of Chinese Academy of Sciences, Beijing 100049, China
College of Electronic and Information Engineering, Southwest University, Chongqing 400715, China
School of Software Engineering, University of Science and Technology of China, Hefei 230051, China
Department of Computer Science, The University of Hong Kong, Hong Kong 999077, China
School of Computer Science, Beijing University of Posts and Telecommunications, Beijing 100876, China

†Huiling Zhang, Min Hao, and Hao Wu contribute equally to this work.

Keywords:
feature extraction, Deep Learning (DL), multi-sequence alignment, residue-residue contact prediction, statistical information, high computational efficiency
Cite this article:
Zhang H, Hao M, Wu H, et al. Protein Residue Contact Prediction Based on Deep Learning and Massive Statistical Features from Multi-Sequence Alignment. Tsinghua Science and Technology, 2022, 27(5): 843-854. https://doi.org/10.26599/TST.2021.9010064
Download citation
EndNote(RIS)
BibTeX

1775

Views

206

Downloads

Citations

8

Crossref

7

WoS

8

Scopus

0

CSCD

Abstract Full text About this article

Sequence-based protein tertiary structure prediction is of fundamental importance because the function of a protein ultimately depends on its 3D structure. An accurate residue-residue contact map is one of the essential elements for current ab initio prediction protocols of 3D structure prediction. Recently, with the combination of deep learning and direct coupling techniques, the performance of residue contact prediction has achieved significant progress. However, a considerable number of current Deep-Learning (DL)-based prediction methods are usually time-consuming, mainly because they rely on different categories of data types and third-party programs. In this research, we transformed the complex biological problem into a pure computational problem through statistics and artificial intelligence. We have accordingly proposed a feature extraction method to obtain various categories of statistical information from only the multi-sequence alignment, followed by training a DL model for residue-residue contact prediction based on the massive statistical information. The proposed method is robust in terms of different test sets, showed high reliability on model confidence score, could obtain high computational efficiency and achieve comparable prediction precisions with DL methods that relying on multi-source inputs.


menu
Abstract
Full text
Outline
About this article

Protein Residue Contact Prediction Based on Deep Learning and Massive Statistical Features from Multi-Sequence Alignment

Show Author's information Huiling Zhang†, Min Hao†, Hao Wu†, Hing-Fung Ting, Yihong Tang( ), Wenhui Xi( ), Yanjie Wei( )
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China
University of Chinese Academy of Sciences, Beijing 100049, China
College of Electronic and Information Engineering, Southwest University, Chongqing 400715, China
School of Software Engineering, University of Science and Technology of China, Hefei 230051, China
Department of Computer Science, The University of Hong Kong, Hong Kong 999077, China
School of Computer Science, Beijing University of Posts and Telecommunications, Beijing 100876, China

†Huiling Zhang, Min Hao, and Hao Wu contribute equally to this work.

Abstract

Sequence-based protein tertiary structure prediction is of fundamental importance because the function of a protein ultimately depends on its 3D structure. An accurate residue-residue contact map is one of the essential elements for current ab initio prediction protocols of 3D structure prediction. Recently, with the combination of deep learning and direct coupling techniques, the performance of residue contact prediction has achieved significant progress. However, a considerable number of current Deep-Learning (DL)-based prediction methods are usually time-consuming, mainly because they rely on different categories of data types and third-party programs. In this research, we transformed the complex biological problem into a pure computational problem through statistics and artificial intelligence. We have accordingly proposed a feature extraction method to obtain various categories of statistical information from only the multi-sequence alignment, followed by training a DL model for residue-residue contact prediction based on the massive statistical information. The proposed method is robust in terms of different test sets, showed high reliability on model confidence score, could obtain high computational efficiency and achieve comparable prediction precisions with DL methods that relying on multi-source inputs.

Keywords: feature extraction, Deep Learning (DL), multi-sequence alignment, residue-residue contact prediction, statistical information, high computational efficiency

References(46)

[1]
J. S. Zhang, W. K. Li, M. Zeng, X. M. Meng, L. Kurgan, F. X. Wu, and M. Li, NetEPD: A network-based essential protein discovery platform, Tsinghua Science and Technology, vol. 25, no. 4, pp. 542–552, 2020.
DOI Google Scholar
[2]
D. S. Marks, T. A. Hopf, and C. Sander, Protein structure prediction from sequence variation, Nat. Biotechnol., vol. 30, no. 11, pp. 1072–1080, 2012.
DOI Google Scholar
[3]
B. Adhikari, D. Bhattacharya, R. Z. Cao, and J. L. Cheng, CONFOLD: Residue-residue contact-guided ab initio protein folding, Proteins: Struct., Funct., Bioinformatics, vol. 83, no. 8, pp. 1436–1449, 2015.
DOI Google Scholar
[4]
J. B. Xu, Distance-based protein folding powered by deep learning, Proc. Natl. Acad. Sci. USA, vol. 116, no. 34, pp. 16856–16865, 2019.
DOI Google Scholar
[5]
A. W. Senior, R. Evans, J. Jumper, J. Kirkpatrick, L. Sifre, T. Green, C. L. Qin, A. Žídek, A. W. R. Nelson, A. Bridgland, H. Penedones, et al., Improved protein structure prediction using potentials from deep learning, Nature, vol. 577, no. 7792, pp. 706–710, 2020.
DOI Google Scholar
[6]
J. Y. Yang, I. Anishchenko, H. Park, Z. L. Peng, S. Ovchinnikov, and D. Baker, Improved protein structure prediction using predicted interresidue orientations, Proc. Natl. Acad. Sci. USA, vol. 117, no. 3, pp. 1496–1503, 2020.
DOI Google Scholar
[7]
M. Baek, F. Dimaio, I. Anishchenko, J. Dauparas, S. Ovchinnikov, G. R. Lee, J. Wang, Q. Cong, L. N. Kinch, R. D. Schaeffer, et al., Accurate prediction of protein structures and interactions using a three-track neural network, Science, vol. 373, no. 6557, pp. 871–876, 2021.
DOI Google Scholar
[8]
A. Raval, S. Piana, M. P. Eastwood, and D. E. Shaw, Assessment of the utility of contact-based restraints in accelerating the prediction of protein structure using molecular dynamics simulations, Protein Sci., vol. 25, no. 1, pp. 19–29, 2016.
DOI Google Scholar
[9]
E. A. Lubecka and A. Liwo, Introduction of a bounded penalty function in contact-assisted simulations of protein structures to omit false restraints, J. Comput. Chem., vol. 40, no. 25, pp. 2164–2178, 2019.
DOI Google Scholar
[10]
Q. Cong, I. Anishchenko, S. Ovchinnikov, and D. Baker, Protein interaction networks revealed by proteome coevolution, Science, vol. 365, no. 6449, pp. 185–189, 2019.
DOI Google Scholar
[11]
D. D. Pollock and W. R. Taylor, Effectiveness of correlation analysis in identifying protein residues undergoing correlated evolution, Protein Eng. Des. Sel., vol. 10, no. 6, pp. 647–657, 1997.
DOI Google Scholar
[12]
S. D. Dunn, L. M. Wahl, and G. B. Gloor, Mutual information without the influence of phylogeny or entropy dramatically improves residue contact prediction, Bioinformatics, vol. 24, no. 3, pp. 333–340, 2007.
DOI Google Scholar
[13]
B. C. Lee and D. Kim, A new method for revealing correlated mutations under the structural and functional constraints in proteins, Bioinformatics, vol. 25, no. 19, pp. 2506–2513, 2009.
DOI Google Scholar
[14]
R. Rajgaria, S. R. McAllister, and C. A. Floudas, Towards accurate residue-residue hydrophobic contact prediction for α helical proteins via integer linear optimization, Proteins: Struct., Funct., Bioinformatics, vol. 74, no. 4, pp. 929–947, 2009.
DOI Google Scholar
[15]
R. Rajgaria, Y. Wei, and C. A. Floudas, Contact prediction for beta and alpha-beta proteins using integer linear optimization and its impact on the first principles 3D structure prediction method ASTRO-FOLD, Proteins: Struct., Funct., Bioinformatics, vol. 78, no. 8, pp. 1825–1846, 2010.
DOI Google Scholar
[16]
J. L. Cheng and P. Baldi, Improved residue contact prediction using support vector machines and a large feature set, BMC Bioinformatics, vol. 8, no. 1, p. 113, 2007.
DOI Google Scholar
[17]
A. N. Tegge, Z. Wang, J. Eickholt, and J. L. Cheng, NNcon: Improved protein contact map prediction using 2D-recursive neural networks, Nucl. Acids Res., vol. 37, no. S2, pp. W515–W518, 2009.
DOI Google Scholar
[18]
S. T. Wu and Y. Zhang, A comprehensive assessment of sequence-based and template-based methods for protein contact prediction, Bioinformatics, vol. 24, no. 7, pp. 924–931, 2008.
DOI Google Scholar
[19]
Z. Y. Wang and J. B. Xu, Predicting protein contact map using evolutionary and physical constraints by integer programming, Bioinformatics, vol. 29, no. 13, pp. i266–i273, 2013.
DOI Google Scholar
[20]
H. L. Zhang, Q. S. Huang, Z. D. Bei, Y. J. Wei, and C. A. Floudas, COMSAT: Residue contact prediction of transmembrane proteins based on support vector machines and mixed integer linear programming, Proteins: Struct., Funct., Bioinformatics, vol. 84, no. 3, pp. 332–348, 2016.
DOI Google Scholar
[21]
M. Weigt, R. A. White, H. Szurmant, J. A. Hoch, and T. Hwa, Identification of direct residue contacts in protein-protein interaction by message passing, Proc. Natl. Acad. Sci. USA, vol. 106, no. 1, pp. 67–72, 2009.
DOI Google Scholar
[22]
D. T. Jones, D. W. A. Buchan, D. Cozzetto, and M. Pontil, PSICOV: Precise structural contact prediction using sparse inverse covariance estimation on large multiple sequence alignments, Bioinformatics, vol. 28, no. 2, pp. 184–90, 2012.
DOI Google Scholar
[23]
F. Morcos, A. Pagnani, B. Lunt, A. Bertolino, D. S. Marks, C. Sander, R. Zecchina, J. N. Onuchic, T. Hwa, and M. Weigt, Direct-coupling analysis of residue coevolution captures native contacts across many protein families, Proc. Natl. Acad. Sci. USA, vol. 108, no. 49, pp. E1293–E1301, 2011.
DOI Google Scholar
[24]
C. Baldassi, M. Zamparo, C. Feinauer, A. Procaccini, R. Zecchina, M. Weigt, and A. Pagnani, Fast and accurate multivariate Gaussian modeling of protein families: Predicting residue contacts and protein-interaction partners, PLoS One, vol. 9, no. 3, p. e92721, 2014.
DOI Google Scholar
[25]
M. Ekeberg, C. Lövkvist, Y. H. Lan, M. Weigt, and E. Aurell, Improved contact prediction in proteins: Using pseudolikelihoods to infer Potts models, Phys. Rev.E, vol. 87, no. 1, p. 012707, 2013.
DOI Google Scholar
[26]
H. Kamisetty, S. Ovchinnikov, and D. Baker, Assessing the utility of coevolution-based residue-residue contact predictions in a sequence-and structure-rich era, Proc. Natl. Acad. Sci. USA, vol. 110, no. 39, pp. 15674–15679, 2013.
DOI Google Scholar
[27]
S. Seemayer, M. Gruber, and J. Söding, CCMpred-fast and precise prediction of protein residue-residue contacts from correlated mutations, Bioinformatics, vol. 30, no. 21, pp. 3128–3130, 2014.
DOI Google Scholar
[28]
M. J. Skwark, A. Abdel-Rehim, and A. Elofsson, PconsC: Combination of direct information methods and alignments improves contact prediction, Bioinformatics, vol. 29, no. 14, pp. 1815–1816, 2013.
DOI Google Scholar
[29]
D. T. Jones, T. Singh, T. Kosciolek, and S. Tetchner., MetaPSICOV: Combining coevolution methods for accurate prediction of contacts and long range hydrogen bonding in proteins, Bioinformatics, vol. 31, no. 7, pp. 999–1006, 2015.
DOI Google Scholar
[30]
B. He, S. M. Mortuza, Y. T. Wang, H. B. Shen, and Y. Zhang, NeBcon: Protein contact map prediction using neural network training coupled with naïve Bayes classifiers, Bioinformatics, vol. 33, no. 15, pp. 2296–2306, 2017.
DOI Google Scholar
[31]
D. T. Jones and S. M. Kandathil, High precision in protein contact prediction using fully convolutional neural networks and minimal sequence features, Bioinformatics, vol. 34, no. 19, pp. 3308–3315, 2018.
DOI Google Scholar
[32]
S. Wang, S. Q. Sun, Z. Li, R. Y. Zhang, and J. B. Xu, Accurate de novo prediction of protein contact map by ultra-deep learning model, PLoS Comput. Biol., vol. 13, no. 1, p. e1005324, 2017.
DOI Google Scholar
[33]
W. Z. Ding, W. Z. Mao, D. Shao, W. X. Zhang, and H. P. Gong, DeepConPred2: An improved method for the prediction of protein residue contacts, Comput. Struct. Biotechnol. J., vol. 16. pp. 503–510, 2018.
DOI Google Scholar
[34]
B. Adhikari, J. Hou, and J. L. Cheng, DNCON2: Improved protein contact prediction using two-level deep convolutional neural networks, Bioinformatics, vol. 34, no. 9, pp. 1466–1472, 2018.
DOI Google Scholar
[35]
B. Adhikari, DEEPCON: Protein contact prediction using dilated convolutional neural networks with dropout, Bioinformatics, vol. 36, no. 2, pp. 470–477, 2020.
DOI Google Scholar
[36]
J. Hanson, K. Paliwal, T. Litfin, Y. D. Yang, and Y. Q. Zhou, Accurate prediction of protein contact maps by coupling residual two-dimensional bidirectional long short-term memory with convolutional neural networks, Bioinformatics, vol. 34, no. 23, pp. 4039–4045, 2018.
DOI Google Scholar
[37]
Q. Wu, Z. L. Peng, I. Anishchenko, Q. Cong, D. Baker, and J. Y. Yang, Protein contact prediction using metagenome sequence data and residual neural networks, Bioinformatics, vol. 36, no. 1, pp. 41–48, 2020.
DOI Google Scholar
[38]
A. Lo, Y. Y. Chiu, E. A. Rødland, P. C. Lyu, T. Y. Sung, and W. L. Hsu, Predicting helix-helix interactions from residue contacts in membrane proteins, Bioinformatics, vol. 25, no. 8, pp. 996–1003, 2009.
DOI Google Scholar
[39]
T. Nugent and D. T. Jones, Predicting transmembrane helix packing arrangements using residue contacts and a force-directed algorithm, PLoS Comput. Biol., vol. 6, no. 3, p. e1000714, 2010.
DOI Google Scholar
[40]
H. L. Zhang, Z. D. Bei, W. H. Xi, M. Hao, Z. Ju, K. M. Saravanan, H. P. Zhang, N. Guo, and Y. J. Wei, Evaluation of residue-residue contact prediction methods: From retrospective to prospective, PLoS Comput. Biol., vol. 17, no. 5, p. e1009027, 2021.
DOI Google Scholar
[41]
D. Kozma, I. Simon, and G. E. Tusnády, PDBTM: Protein data bank of transmembrane proteins after 8 years, Nucl. Acids Res., vol. 41, no. D1, pp. D524–D529, 2013.
DOI Google Scholar
[42]
Y. Zhang, J. W. T. Chan, F. Y. L. Chin, H. F. Ting, D. S. Ye, F. Zhang, and J. Y. Shi, Constrained pairwise and center-star sequences alignment problems, J. Comb. Optim., vol. 32, no. 1, pp. 79–94, 2016.
DOI Google Scholar
[43]
W. T. Chan, Y. Zhang, S. P. Y. Fung, D. S. Ye, and H. Zhu, Efficient algorithms for finding a longest common increasing subsequence, J. Comb. Optim., vol. 13, no. 3, pp. 277–288, 2007.
DOI Google Scholar
[44]
C. X. Zhang, W. Zheng, S. M. Mortuza, Y. Li, and Y. Zhang, DeepMSA: Constructing deep multiple sequence alignment to improve contact prediction and fold-recognition for distant-homology proteins, Bioinformatics, vol. 36, no. 7, pp. 2105–2112, 2020.
DOI Google Scholar
[45]
A. J. Hockenberry and C. O. Wilke, Evolutionary couplings detect side-chain interactions, PeerJ, vol. 7, p. e7280, 2019.
DOI Google Scholar
[46]
M. Chonofsky, S. H. P. De Oliveira, K. Krawczyk, and C. M. Deane, The evolution of contact prediction: Evidence that contact selection in statistical contact prediction is changing, Bioinformatics, vol. 36, no. 6, pp. 1750–1756, 2020.
DOI Google Scholar
Publication history
Copyright
Acknowledgements
Rights and permissions

Publication history

Received: 20 July 2021
Revised: 17 August 2021
Accepted: 20 August 2021
Published: 17 March 2022
Issue date: October 2022

Copyright

© The author(s) 2022.

Acknowledgements

This work was partly supported by the Strategic Priority CAS Project (No. XDB38050100), the National Key Research and Development Program of China (No. 2018YFB0204403), the National Natural Science Foundation of China (No. U1813203), the Shenzhen Basic Research Fund (Nos. RCYX2020071411473419, JCYJ20200109114818703, and JSGG20201102163800001), CAS Key Lab (No. 2011DP173015), Hong Kong Research Grant Council (No. GRF-17208019), and the Outstanding Youth Innovation Fund (Doctoral Students) of CAS-SIAT (No. Y9G054).

Rights and permissions

The articles published in this open access journal are distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/).

Return