Affiliation:
1. I. Sechenov Institute of Evolutionary Physiology and Biochemistry, Russian Academy Sciences
Abstract
Relevance of searching for computer methods with high efficiency of immunoepitopes recognition and predicting the longevity of the immunity they induce is determined primarily by the need to quickly create vaccines against newly emerging infections, especially during pandemic periods. Aim. To develop a new immunoinformation method for recognizing immunoepitopes, to identify in the viral proteins possible potential markers to induce long-term immunity and to evaluate by them the vaccines against Covid-19. Materials and methods. For computer analysis, an Internet-accessible databases of immunoep topes 15 and 9 amino acids long, restricted respectively by MHC I and MHC II, and peptides not binding to MHC, as well as human and virus proteins, were used. The algorithm for discriminating immunoepitopes was based on positional distinction of specific short peptides in their primary structures. Results. The «inventory» in the training samples of di- and tripeptides or pentapeptides of immunoepitopes and nonimmunoepitopes makes it possible to accurately recognize in the control samples up to 93–97% of immunoepitopes restricted by MHC I and MHC II. Comparison of the amino acid composition of proteins of subunit vaccines causing long-term immunity revealed dominance of amino acids (especially proline), which form the basis of internally disorganized regions, and proline-containing dipeptides, that allowed them to be considered as biomarkers of the potential of a viral protein to form a long-term immune memory.In the S-protein of coronavirus SARS-CoV-2 two candidates for immunospressive domains are present and the dominance of proline and dipeptides containing it is absent. Conclusion. The immunoepitope recognition method and the biomarker for inducing longterm immune memory can be used as immunoinformative tools of computational vaccinology. Providing long-term immunity by vaccines based on the coronavirus SARS-CoV-2 protein S is unlikely.
Subject
Infectious Diseases,Public Health, Environmental and Occupational Health,Epidemiology
Reference23 articles.
1. Joglekar AV, Li G. T cell antigen discovery. Nature Methods. 2020. Doi:10.1038/s41592-020-0867-z
2. Peters B, Nielsen M, Sette A. T Cell Epitope Predictions. Annu. Rev. Immunol. 2020. 38:123–45. Doi 10.1146/annurev-immunol-082119-124838
3. Loan Ping Eng, Tin Wee Tan, Joo Chuan Tong , Söllner J. Building MHC Class II Epitope Predictor Using Machine Learning Approaches. In: Peng Zhou and Jian Huang (eds.), Computational Peptidology, Methods in Molecular Biology, vol. 1268, DOI 10.1007/978-1-4939-2285-7_4,
4. Söllner J. Computational Peptide Vaccinology. In: Peng Zhou and Jian Huang (eds.), Computational Peptidology, Methods in Molecular Biology, vol. 1268, DOI: 10.1007/978-1-4939-2285-7_13
5. Prigozhin I.R., Stengers I. Order from chaos. Progress 1986. s432.
Cited by
5 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献