informИнформатикаInformatics1816-03012617-6963UIIP NASB10.37661/1816-0301-2020-17-2-44-53inform-1068Research ArticleБИОИНФОРМАТИКАBIOINFORMATICSПредсказание структуры гомодимерных белковых комплексов на основе глубокой нейронной сетиProtein homodimers structure prediction based on deep neural networkХадаровичА. Ю.HadarovichA. Y.

Хадарович Анна Юрьевна, научный сотрудник

Минск

Anna Y. Hadarovich, Researcher

Minsk

ahadarovich@gmail.comКалиновскийА. А.KalinouskiA. A.

Калиновский Александр Александрович, научный сотрудник

Минск

Alexander A. Kalinouski, Researcher

Minsk

gakarak@gmail.comТузиковА. В.TuzikovA. V.

Тузиков Александр Васильевич, член-корреспондент, доктор физико-математических наук, профессор, генеральный директор

Минск

Alexander V. Tuzikov, Corresponding Member, Dr. Sci. (Phys.-Math.), Professor, General Director

Minsk

tuzikov@newman.bas-net.byОбъединенный институт проблем информатики Национальной академии наук Беларуси; Белорусский государственный университетThe United Institute of Informatics Problems of the National Academy of Sciences of Belarus; Belarusian State UniversityОбъединенный институт проблем информатики Национальной академии наук БеларусиThe United Institute of Informatics Problems of the National Academy of Sciences of Belarus2020070520201724453Copyright © Хадарович А.Ю., Калиновский А.А., Тузиков А.В., 20202020Хадарович А.Ю., Калиновский А.А., Тузиков А.В.Hadarovich A.Y., Kalinouski A.A., Tuzikov A.V.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://inf.grid.by/jour/article/view/1068

Предсказание структур белковых комплексов имеет важные приложения в таких областях, как моделирование биологических процессов и разработка лекарственных средств. Гомодимеры (комплексы, состоящие из двух одинаковых белков) являются наиболее распространенным типом белковых комплексов в природе, но до сих пор нет универсального алгоритма для предсказания их трехмерных структур. Экспериментальные методы для определения структур белковых комплексов требуют значительных затрат времени и ресурсов и имеют свои ограничения. Глубокие нейронные сети позволили предсказать структуры отдельных белков, значительно превзойдя по точности другие алгоритмические подходы. Опираясь на идею этого подхода, в статье разработан алгоритм для моделирования трехмерной структуры гомодимерных комплексов на основе глубокого обучения. Он состоит из двух основных этапов: на первом этапе прогнозируется карта контактов белкового комплекса при помощи глубокой сверточной нейронной сети, а на втором предсказывается трехмерная структура гомодимера на основе полученной карты контактов и процедуры оптимизации. Предложенный подход был протестирован и проверен на наборе белковых гомодимеров из базы данных белков PDB (Protein Data Bank). Разработанная методика может быть использована для оценки моделей белковых гомодимеров в качестве одного из этапов разработки лекарственных соединений.

Structural prediction of protein-protein complexes has important application in such domains as modeling of biological processes and drug design. Homodimers (complexes which consist of two identical proteins) are the most common type of protein complexes in nature but there is still no universal algorithm to predict their 3D structures. Experimental techniques to identify the structure of protein complex require enormous amount of time and resources, and each method has its own limitations. Recently Deep Neural Networks allowed to predict structures of individual proteins greatly prevailing in accuracy over other algorithmic approaches. Building on the idea of this approach, we developed an algorithm to model the 3D structure of homodimer based on deep learning. It consists of two major steps: at the first step a protein complex contact map is predicted with the deep convolutional neural network, and the second stage is used to predict 3D structure of homodimer based on obtained contact map and optimization procedure. The use of the neural network in combination with optimization procedure based on gradient descent method allowed to predict structures for protein homodimers. The suggested approach was tested and validated on a dataset of protein homodimers from Protein Data Bank (PDB). The developed procedure could be also used for evaluating protein homodimer models as one of the stages in drug compounds developing.

методы глубокого обучениянейронные сетимоделирование белковых комплексовгомодимерыоптимизацияdeep learningneural networkmodeling of protein complexeshomodimersoptimizationReferencesAnfinsen C. B. Principles that govern the folding of protein chains. Science, 1973, vol. 181 (4096), pp. 223–230. https://doi.org/10.1126/science.181.4096.223Anfinsen C. B. Principles that govern the folding of protein chains. Science, 1973, vol. 181 (4096), pp. 223–230. https://doi.org/10.1126/science.181.4096.223Lecun Y., Bengio Y., Hinton G. Deep learning. Nature, 2015, vol. 521 (7553), pp. 436–444. https://doi.org/10.1038/nature14539Lecun Y., Bengio Y., Hinton G. Deep learning. Nature, 2015, vol. 521 (7553), pp. 436–444. https://doi.org/10.1038/nature14539Senior A. W., Evans R., Jumper J., Kirkpatrick J., Sifre L., …, Hassabis D. Improved protein structure prediction using potentials from deep learning. Nature, 2020, vol. 577 (7792), pp. 706–710. https://doi.org/10.1038/s41586-019-1923-7Senior A. W., Evans R., Jumper J., Kirkpatrick J., Sifre L., …, Hassabis D. Improved protein structure prediction using potentials from deep learning. Nature, 2020, vol. 577 (7792), pp. 706–710. https://doi.org/10.1038/s41586-019-1923-7Billings W. M., Hedelius B., Millecam T., Wingate D., Corte D. D. ProSPr: democratized implementation of alphafold protein distance prediction network. Biorxiv, 2019, p. 830273. https://doi.org/10.1101/830273Billings W. M., Hedelius B., Millecam T., Wingate D., Corte D. D. ProSPr: democratized implementation of alphafold protein distance prediction network. Biorxiv, 2019, p. 830273. https://doi.org/10.1101/830273Kryshtafovych A. T., Schwede, Topf M., Fidelis K., Moult J. Critical assessment of methods of protein structure prediction (CASP) – Round XIII. Proteins: Structure, Function, and Bioinformatics, 2019, vol. 87 (12), pp. 1011–1020. https://doi.org/10.1002/prot.25823Kryshtafovych A. T., Schwede, Topf M., Fidelis K., Moult J. Critical assessment of methods of protein structure prediction (CASP) – Round XIII. Proteins: Structure, Function, and Bioinformatics, 2019, vol. 87 (12), pp. 1011–1020. https://doi.org/10.1002/prot.25823Jones D. T., Kandathil S. M. High precision in protein contact prediction using fully convolutional neural networks and minimal sequence features. Bioinformatics, 2018, vol. 34, pp. 3308–3315.Jones D. T., Kandathil S. M. High precision in protein contact prediction using fully convolutional neural networks and minimal sequence features. Bioinformatics, 2018, vol. 34, pp. 3308–3315.Seemayer S., Gruber M., Söding J. CCMpred – fast and precise prediction of protein residue-residue contacts from correlated mutations. Bioinformatics, 2014, vol. 30 (21), pp. 3128–3130.Seemayer S., Gruber M., Söding J. CCMpred – fast and precise prediction of protein residue-residue contacts from correlated mutations. Bioinformatics, 2014, vol. 30 (21), pp. 3128–3130.Jones D. T., Singh T., Kosciolek T., Tetchner S. MetaPSICOV: combining coevolution methods for accurate prediction of contacts and long range hydrogen bonding in proteins. Bioinformatics, 2015, vol. 31 (7), pp. 999–1006.Jones D. T., Singh T., Kosciolek T., Tetchner S. MetaPSICOV: combining coevolution methods for accurate prediction of contacts and long range hydrogen bonding in proteins. Bioinformatics, 2015, vol. 31 (7), pp. 999–1006.Jiang Q., Jin X., Lee S.-J., Yao S. Protein secondary structure prediction: a survey of the state of the art. Journal of Molecular Graphics and Modelling, 2017, vol. 76, pp. 379–402. https://doi.org/10.1016/j.jmgm.2017.07.015Jiang Q., Jin X., Lee S.-J., Yao S. Protein secondary structure prediction: a survey of the state of the art. Journal of Molecular Graphics and Modelling, 2017, vol. 76, pp. 379–402. https://doi.org/10.1016/j.jmgm.2017.07.015Skwark M. J., Raimondi D., Michel M., Elofsson A. Improved contact predictions using the recognition of protein like contact patterns. PLoS Computational Biology, 2014, vol. 10 (11), p. e1003889. https://doi.org/10.1371/journal.pcbi.1003889Skwark M. J., Raimondi D., Michel M., Elofsson A. Improved contact predictions using the recognition of protein like contact patterns. PLoS Computational Biology, 2014, vol. 10 (11), p. e1003889. https://doi.org/10.1371/journal.pcbi.1003889Berman H. M. The protein data bank: a historical perspective. Acta Crystallographica Section A: Foundations of Crystallography, 2008, vol. 64 (1), pp. 88–95. https://doi.org/10.1107/S0108767307035623Berman H. M. The protein data bank: a historical perspective. Acta Crystallographica Section A: Foundations of Crystallography, 2008, vol. 64 (1), pp. 88–95. https://doi.org/10.1107/S0108767307035623Mou Y., Huang P.-S., Hsu F.-C., Huang S.-J., Mayo S. L. Computational design and experimental verification of a symmetric protein homodimer. Proceedings of the National Academy of Sciences of the United States of America, 2015, vol. 112 (34), pp. 10714–10719. https://doi.org/10.1073/pnas.1505072112Mou Y., Huang P.-S., Hsu F.-C., Huang S.-J., Mayo S. L. Computational design and experimental verification of a symmetric protein homodimer. Proceedings of the National Academy of Sciences of the United States of America, 2015, vol. 112 (34), pp. 10714–10719. https://doi.org/10.1073/pnas.1505072112Long J., Shelhamer E., Darrell T. Fully convolutional networks for semantic segmentation. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Boston, Massachusetts, USA, 7–12 June 2015. Boston, 2015, pp. 3431–3440. https://doi.org/10.1109/CVPR.2015.7298965Long J., Shelhamer E., Darrell T. Fully convolutional networks for semantic segmentation. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Boston, Massachusetts, USA, 7–12 June 2015. Boston, 2015, pp. 3431–3440. https://doi.org/10.1109/CVPR.2015.7298965Fu J., Liu J., Tian H., Li Y., Bao Y., …, Lu H. Dual attention network for scene segmentation. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Long Beach, California, USA, 15–20 June 2019. Long Beach, 2019, pp. 3141–3149. Available at: http://arxiv.org/abs/1809.02983 (accessed 27.03.2020).Fu J., Liu J., Tian H., Li Y., Bao Y., …, Lu H. Dual attention network for scene segmentation. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Long Beach, California, USA, 15–20 June 2019. Long Beach, 2019, pp. 3141–3149. Available at: http://arxiv.org/abs/1809.02983 (accessed 27.03.2020).He K., Zhang X., Ren S., Sun J. Deep residual learning for image recognition. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Las Vegas, Nevada, USA, 27–30 June 2016. Las Vegas, 2016, pp. 770–778. https://doi.org/10.1109/CVPR.2016.90He K., Zhang X., Ren S., Sun J. Deep residual learning for image recognition. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Las Vegas, Nevada, USA, 27–30 June 2016. Las Vegas, 2016, pp. 770–778. https://doi.org/10.1109/CVPR.2016.90Chen L.-C., Papandreou G., Kokkinos I., Murphy K., Yuille A. L. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, vol. 40 (4), pp. 834–848. https://doi.org/10.1109/TPAMI.2017.2699184Chen L.-C., Papandreou G., Kokkinos I., Murphy K., Yuille A. L. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, vol. 40 (4), pp. 834–848. https://doi.org/10.1109/TPAMI.2017.2699184Kingma D. P., Ba J. Adam: a method for stochastic optimization. 3rd International Conference on Learning Representations (ICLR 2015), San Diego, CA, USA, 7–9 May 2015. San Diego, 2015. Available at: https://arxiv.org/abs/1412.6980 (accessed 27.03.2020).Kingma D. P., Ba J. Adam: a method for stochastic optimization. 3rd International Conference on Learning Representations (ICLR 2015), San Diego, CA, USA, 7–9 May 2015. San Diego, 2015. Available at: https://arxiv.org/abs/1412.6980 (accessed 27.03.2020).Mitternacht S. FreeSASA: an open source C library for solvent accessible surface area calculations. F1000Research, 2016, vol. 5, p. 189. https://doi.org/10.12688/f1000research.7931.1Mitternacht S. FreeSASA: an open source C library for solvent accessible surface area calculations. F1000Research, 2016, vol. 5, p. 189. https://doi.org/10.12688/f1000research.7931.1Janin J., Bahadur R. P., Chakrabarti P. Protein-protein interaction and quaternary structure. Quarterly Reviews of Biophysics, 2008, vol. 41 (2), pp. 133–180. https://doi.org/10.1017/S0033583508004708Janin J., Bahadur R. P., Chakrabarti P. Protein-protein interaction and quaternary structure. Quarterly Reviews of Biophysics, 2008, vol. 41 (2), pp. 133–180. https://doi.org/10.1017/S0033583508004708Cock P. J., Antao T., Chang J. T., Chapman B. A., Cox C. J., …, de Hoon M. J. Biopython: freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics, 2009, vol. 25 (11), pp. 1422–1423. https://doi.org/10.1093/bioinformatics/btp163Cock P. J., Antao T., Chang J. T., Chapman B. A., Cox C. J., …, de Hoon M. J. Biopython: freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics, 2009, vol. 25 (11), pp. 1422–1423. https://doi.org/10.1093/bioinformatics/btp163

The authors declare that there are no conflicts of interest present.