Biomedical and Biotechnology Research Journal (BBRJ)

: 2022  |  Volume : 6  |  Issue : 3  |  Page : 319--325

Immunoinformatics-based identification of highly conserved cytotoxic T-cell epitopes in polyprotein pp220 of african swine fever virus

Aiman Kiara Atienza Juan, Keana Milen Calara Palma, Marianne Bermudez Suarez, Leana Rich De Mesa Herrera-Ong 
 Department of Physical Sciences, College of Science, Polytechnic University of the Philippines, Manila, Philippines

Correspondence Address:
Aiman Kiara Atienza Juan
A. Mabini Campus, Anonas St., Sta. Mesa Manila, 1016 Metro Manila


Background: High mortality rate of pigs peaked in 2020 due to the re-emergence of a deadly African swine fever virus (ASFV) which has led to transcontinental outbreaks in Europe, reportedly from 2014 to 2019, and in Asia and the Pacific from 2018–2020. Given the huge socioeconomic consequences of the disease, vaccines that will prime the immunity of swine against this pathogen is a dire necessity. Methods: In silico identification and characterization of highly conserved cytotoxic T-cell (CD8+) epitopes derived from one of its structural proteins, pp220, were analyzed. Protein sequences of pp220 were retrieved and clustered to obtain highly conserved sequences. Cross-reactive epitopes were filtered out, and the remaining epitopes were docked with swine leukocyte antigen-1*0401 (SLA-1*0401). Furthermore, the epitope stability was determined by comparing binding energy, dissociation constant, and eigenvalues of the epitopes with the values of positive control, influenza-epitope complex. Results: This study showed that 20 highly conserved epitopes promiscuously bind to two or more SLAs and 9 of which epitopes (ALDLSLIGF, QIYKTLLEY, FLNKSTQAY, IADAINQEF, IINPSITEY, AINTFMYYY, SLYPTQFDY, RSNPGSFYW, and RLDRKHILM) that were validated exhibit potential immunogenicity based on the acceptable binding energy, dissociation constant, and eigenvalues. Conclusion: This study has identified epitopes that show high conservancy, reducing the chance of epitope immune evasion. It is anticipated that the identified epitopes must be further evaluated as a potential immunotherapeutic agent in developing an epitope-based vaccine against ASFV.

How to cite this article:
Juan AK, Palma KM, Suarez MB, Herrera-Ong LR. Immunoinformatics-based identification of highly conserved cytotoxic T-cell epitopes in polyprotein pp220 of african swine fever virus.Biomed Biotechnol Res J 2022;6:319-325

How to cite this URL:
Juan AK, Palma KM, Suarez MB, Herrera-Ong LR. Immunoinformatics-based identification of highly conserved cytotoxic T-cell epitopes in polyprotein pp220 of african swine fever virus. Biomed Biotechnol Res J [serial online] 2022 [cited 2022 Nov 28 ];6:319-325
Available from:

Full Text


Originating from East Africa, African swine fever (ASF) was identified as a disease with a high fatality rate of 100%, with wild warthogs (Phacochoerus spp) as the suspected primary infection source.[1] It has been traced to re-emerge as it is caused by the high virulent ASF virus (ASFV). This virus infects domestic pigs (Sus scrofa domesticus), warthogs, bush pigs, wild boars, and soft ticks (Ornithodoros erraticus) that possibly mediate the transmission.[2],[3] ASF is identified to be transmissible through direct and indirect contact infection of the host which is attributed to passive vectors such as fomites and human–swine interaction, along with the equipment exposed during pig care and transportation. The ASF virus is present in all swine secretions and excretions and can be lethal depending on the virulence of the determined isolate.[1] It causes dire physiological damages to swine as clinically examined.[4]

Currently, the infection of numerous wild boars and domestic pigs is continuous, and data reports show that transcontinental outbreaks spread through member states of the European Union from 2014 to 2019 while heightened in Asia and the Pacific in 2018–2020.[5],[6],[7],[8],[9] The ASF epidemic has caused adverse effects on the pig trade industry resulting in decreased local and international pork meat supply, thus, lower economic contribution.[7],[8],[9] Efforts to establish long-time control and prevention of the ASF disease aim for immediate vaccine development alongside the stringent measures depending on the virus case proliferation conduct that varies from one local place to another.[6] Multiple experimental techniques are ongoing to determine how to mitigate the proliferation of the infection, but the heightened restriction is the only treatment there is to maintain high biosecurity, prevent further infections, and protect potential hosts since no commercially licensed vaccine is available up to date, but several subunits, DNA, and virus vector vaccine studies are being explored.[1],[10]

The ASFV is determined as a cytoplasmic double-stranded DNA virus. ASFV encoded structural proteins (SPs) perform vital roles in viral infection as determined in 11 complete ASFV genome sequences representing different isolates.[3] Among the SPs, pp220 is one of the proteins that influence the assembly of mature virions such that coreless or infective virus particles may be observed when pp220 processing was obstructed. Polyprotein pp220 serves as a protein precursor shortly cleaved to produce several mature virion proteins. It is involved in genome replication and viral infection and performs roles in entry and attachment encoded by the gene CP2475 L.[3],[11] Favorably, gene name, CP2475 L, the same gene encoded the protein of interest, pp220, is one of the pools of DNA used in expressing selected antigens in pig immunization have recorded an immune response after a lethal dose of ASFV isolate Georgia 2007/1 which suggests possible enhancement in immunity of pigs inoculated. Specifically, viral genomes levels decreased observably in blood and some lymph tissues in the pigs immunized.[11] These characteristics show the relevance of polyprotein pp220 as a good protein of interest that can be utilized for epitope-based vaccine development to protect swine from ASFV.

T-cell epitopes bind to major histocompatibility complex (MHC) which then interact with T-cell receptors and eventually induce an immune response.[12] Identification of T-cell epitopes could be done through epitope mapping using bioinformatics tools. Informatics studies provide a useful tool in developing biologic-based therapeutics and are able to give a preassessment of its effectiveness.[13] This bridges the gap between time-consuming and expensive repetitive experimental setup and heavy computational approaches. In silico identification of putative epitopes can potentially aid in the development of subunit vaccines that can provide long-lasting immunity against the virus. With the use of several immunoinformatics tools, this work aims to identify highly conserved CD8+ epitopes against pp220 polyprotein of ASFV. The identified epitopes can be used for several important purposes, including the development of epitope-based vaccines, assessment of physiological disease state, origin-tracing, immunological response-monitoring, and further improvement of existing diagnosis assays.[14]


Identification of highly conserved protein sequences from pp220 polyprotein of African swine fever virus

All pp220 ASFV amino acid sequences with a length range of 2475–2476 were acquired from National Center for Biotechnology Information (NCBI) GenBank ( Moreover, topological information such as glycosylation and cleavage sites was obtained from UniProt and was used as part of the assessment in epitope screening. Acquiring a large number of amino acid sequences allows better identification of the highly conserved sequences. Collected pp220 sequences of ASFV were analyzed using CD-HIT ( cmd = cd-hit) and Clustal Omega ( server tools in determining highly conserved protein sequence. CD-HIT allows the users to generate representative sequences out of the possible repetitively acquired pp220 sequences through clustering function or by comparing each dataset even with varying identity levels.[15] In the case of the preferred sequence identity cutoff, 1.0 is intended, which means a 100% unique identity target, and the rest of the parameters were set on default.

Connectedly, through Clustal Omega, the user can conduct multiple sequence alignments of the representative sequence in text file format coming from the CD-HIT output. Clustal Omega aligns large datasets of protein sequences with accuracy, and it is known to outperform other available same functional tools in terms of quality and timely execution.[16] The obtained representative sequences were uploaded to Clustal Omega with default parameters unchanged. The resulting list of aligned sequences retrieved in CLUSTAL_NUM format was uploaded to the Protein Variability Server (PVS) ( for protein variability analysis. PVS can analyze the variability on-site in multiple protein-sequence alignments essential to identify conserved sequences and produce variability-masked sequences.[17] Initially, the Shannon variability method and variability threshold of 0.1 were selected to obtain residues of highly conserved positions.[18] Highly conserved sequences are identified to exhibit a variability less than or equal to 1.0, implying a lower rate of possible sequence variation while masked sequences, in contrast, are fragments with high sequence variation and thus not highly conserved. Also, the length of the fragment was a minimum of 9 amino acids, sorted according to their position in multiple sequence alignment and the rest of the parameters at default.[17]

Identification of candidate cytotoxic T-cell epitopes

The identified highly conserved sequences of pp220 were assessed for the presence of epitopes with good binding affinities to the 45 swine leukocytes alleles available in the Immune Epitope Database ( tool. More particularly, the process was conducted using the proteasomal cleavage/transporter associated with antigen processing transport (TAP)/MHC I combined predictor tool, along with NetMHCcons algorithm. The algorithm NetMHCcons is a combination of two MHC-peptide binding tools (NetMHCpan and Pickpocket), which outweighs the accuracy of other tools.[19] Peptides with 9 residues, IC50 <500 nmol/dm3, proteasome-processing score >0, TAP score >0, and sequences that are not situated within the determined cleavage sites were selected.[20] Moreover, epitopes that bind to four or more SLA will be considered a “promiscuous” epitopes and will be further analyzed for cross-reactivity. Protein–protein Basic Local Alignment Search Tool (BLASTp) ( PAGE = Proteins) was used to test the possible cross-reactivity of the identified epitopes to available swine protein sequences in UniProt and Swiss-Prot databases. Cross-reactivity occurs when the amino acid sequences of pathogen and self-tissue proteins are similar which may result in autoimmune reactions.[21],[22]

Molecular docking and molecular dynamics simulation of major histocompatibility complex-epitope complexes with the candidate epitopes

Candidate epitopes that passed through the in silico cross-reactivity analysis and bound to SLA-1*0401 (the only swine MHC I with elucidated structure available in the used server used up to date) were subjected to structure validation through docking analysis. The crystal structure of SLA-1*0401 (Protein Data Bank [PDB] 3QQ3) was downloaded from the Research Collaboratory for Structural Bioinformatics PDB (, a database collection of 3D structures of different biomolecules.[23] PDB 3QQ3 is a crystal structure of SLA-1*0401 and is bound to an influenza epitope. The structure was cleaned by removing the bound influenza epitope for subsequent use in molecular docking of candidate cytotoxic T-cell epitopes. The PDB structure of SLA-1*0401 was docked with the epitopes using the GalaxyPepDock ( type = PEPDOCK) from GalaxyWEB server ([24] This server tool is useful for building models and docking by locating templates of experimentally defined structures that can be used to effectively demonstrate the structural differences between the target protein-peptide complex and the template.[25] Retrieved epitope-SLA complex structures from GalaxyPepDock were refined using the GalaxyRefine Complex tool (!) which performs refinement proven to successfully enhance local and global structure accuracy through side-chain repacking, a method tested in Critical Assessment of techniques for protein Structure Prediction 10.[26] Afterward, the epitopes were elucidated using iCn3D (https://www. ncbi. nlm. nih. gov/Structure/icn3d/full. html), an integrated sequence/annotation server tool of NCBI that displays 3D structure, 2D molecular interactions, and 1D protein and nucleotide sequences.[27]

On the other hand, the binding capability of candidate epitopes to SLA-1*0401 was validated using PRODIGY WebServer ([28] PRODIGY is a set of online services aimed to predict binding affinity in biological complexes and distinguish biological interfaces of crystallographic structures.[29] Procured peptide sequence structures were uploaded and analyzed using the PRODIGY protein–protein server, wherein interactors are filled in based on the PDB structure complex's chain identity at a standard temperature of 37°C. Afterward, the obtained results were compared to the values of positive control PDB 3QQ3.

Further evaluation of the binding stability of the candidate epitopes was conducted through molecular dynamics and carried out using iMODs (, a fast, easy, and user-friendly online simulation tool used for the dynamics simulation study. Eigenvalue is a vital parameter for determining the correlation of the residue–residue matrix of interest through molecular dynamics. This value indicates to what extent a specific deformation motion affects the whole structural protein complex. Lower the eigenvalue indicates that the protein complex is more prone to deformation.[30],[31]


Collection and preparation of the ASFV polyprotein pp220

Primarily, the NCBI database provided sequences of pp220 from various strains encoded by the same gene CP2475 L, with an amino acid range count from 2475 to 2476 upon access date. The use of UniProt then identified the cleavage sites specifically positioned at amino acids 45–46, 369–370, 523–524, and 894–895. In the later part of the study, resulting epitopes containing these cleavage sites were removed since these sites are less likely to be part of a functional epitope.[32] Thereafter, running to CD-HIT and Clustal Omega in silico tools generated an exact number of 8 programmatically aligned representative sequences.

Identification of highly conserved candidate cytotoxic T-cell epitopes

Substantially, PVS presented highly conserved protein sequences and their amino acid numbered positions. This produced 69 peptide fragments composed of at least nine consecutive residues, all of which are used in locating the target epitopes, ensuring they are positioned within any of the fragments generated. IEDB predicted 20 cytotoxic T-cell epitopes that promiscuously bind to at least two SLAs; nine were found to bind with SLA-1*0401, as shown in [Table 1]. SLA-1*0401 is one of the most prevalent SLA that can reinforce essential antigen recognition, especially due to the known genetic polymorphism of MHC (SLA) I, thus covering a broader swine population.[33],[34] Each of the peptides has MHC IC50 <500, and its corresponding proteasome and TAP score are positive.{Table 1}


The parameters used to account for possible cross-reactivity are E-value, percent identity, and query coverage which were all obtained using the BLASTp program. The query coverage of the epitopes lies within the range of 66%–100%. These metrics can indicate significant sequence similarity between the epitope and the protein sequences queried in the database.

Upon evaluation, the identified epitopes have an E-value ranging from 0.19-154. SDMPGVQLI has the highest E-value (E-value=154) and SQWDLVQKF has the lowest (E-value=0.19). Moreover, one of the epitopes (RERERIFNL) shows “no significant match” in the database.

Molecular docking and simulation of MHC – epitope complexes

Retrieved epitopes were docked with the refined structure of SLA-1*0401, wherein models with the highest accuracy were chosen as representative MHC-epitope complex structures. [Figure 1] shows the docked structures of (A) ALDLSLIGF, (B) QIYKTLLEY, (C) IADAINQEF, (D) FLNKSTQAY, (E) IINPSITEY, (F) AINTFMYYY, (G) SLYPTQFDY, (H) RSNPGSFYW, and (I) RLDRKHILM. This figure displays that the retrieved candidate epitopes bind to the SLA-1*0401 binding groove. Notably presented in [Table 2] are values of calculated binding energy (ΔG) and dissociation constant (Kd) for each matrix for further evaluation.{Figure 1}{Table 2}

Among the nine epitope candidates, RSNPGSFYW has the closest binding energy and affinity values with positive control. The remaining epitopes show comparable values, indicating favorable MHC-epitope binding formation may occur. On the other hand, the stability of the binding epitopes was determined by calculating the eigenvalues (shown in [Table 3]).{Table 3}

From the obtained epitopes, ALDLSLIGF and SLYPTQFDY obtained the closest eigenvalues from the positive control, whereas the other epitopes had higher eigenvalues which indicate less susceptibility to protein complex deformation.[30]


ASFV has brought an alarming crisis in the poultry industry as well as in the global market. Along with the mitigation plans being implemented to counterattack the repercussions of this pandemic, different vaccine development is also now in the works. Bioinformatics has been used in vaccine design for the past years as it offers cost-effective and time-efficient in silico analysis. In this study, bioinformatics is used along with the concept of epitopes as a potential immunotherapeutic agent against diseases to help the progression of vaccine development. The cytotoxic T-cell epitopes obtained from highly conserved sequences went through in silico processes and were narrowed down to nine epitopes that can be potentially used for vaccine development.

Peptides containing MHC IC50 <500 nM indicate good binding relation. Epitopes that have an E-value greater than 1.0e-30 are less likely to cross-react with swine proteome.[35] In some rare cases, cross-reactivity happens for matches with 50% identity, but in most cases, matches with >70% are more likely to induce in vitro cross-reactivity and/or show clinical relevance.[36] ΔG, Kd, and eigenvalues were used to evaluate epitope-binding tendency and MHC I-epitope complex stability. Favorable epitope binding requires low binding energy and dissociation constant to indicate favorable formation of MHC I-epitope complex. The MHC-epitope docked structures in this study obtained negative ΔG values and dissociation constants that fall within the range of peptide MHC binding Kd values of 10-7-10-8,[37],[38] which means epitopes efficiently binds with MHC binding groove of SLA-1*0401. Moreover, eigenvalues are associated with the motion stiffness of molecules. Its value is proportional to the amount of energy necessary to disfigure the structure; thus, the lower the eigenvalues, the more prone the biomolecule to structure deformation as it requires a lesser amount of energy to be disfigured.[33]


In silico identified cytotoxic T-cell epitopes from ASFV pp220 including ALDLSLIGF, QIYKTLLEY, FLNKSTQAY, IADAINQEF, IINPSITEY, AINTFMYYY, SLYPTQFDY, RSNPGSFYW, and RLDRKHILM demonstrate good binding capability, acceptable stability with MHC-binding groove, and low potential cross-reactivity. Most importantly, candidate epitopes are part of highly conserved protein sequences to avoid immune epitope evasion. It is anticipated that these candidate epitopes will be further evaluated through in vivo or in vitro assays. Overall, these epitopes can potentially aid in the development of vaccines against ASFV.

Limitation of study

This work is only limited to the identification of cytotoxic T-cell epitopes specifically from pp220 of ASFV which is composed of 2475–2476 amino acids, that could help in designing a potential vaccine against ASFV. Therefore, thorough vaccine design is excluded in this study.

Ethical statement

There is no unethical method used in this research.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Spickler, Anna Rovid. African Swine Fever. Africaine PP; June, 2019. Available from: [Last acessed on 2021 Oct 29].
2Sánchez-Cordón PJ, Montoya M, Reis AL, Dixon LK. African swine fever: A re-emerging viral disease threatening the global pig industry. Vet J 2018;233:41-8.
3Jia N, Ou Y, Pejsak Z, Zhang Y, Zhang J. Roles of African swine fever virus structural proteins in viral infection. J Vet Res 2017;61:135-43.
4Pikalo J, Zani L, Hühr J, Beer M, Blome S. Pathogenesis of African swine fever in domestic pigs and European wild boar–Lessons learned from recent animal trials. Virus Res 2019;271:197614.
5Blome S, Franzke K, Beer M. African swine fever-A review of current knowledge. Virus Res 2020;287:198099.
6Dixon LK, Stahl K, Jori F, Vial L, Pfeiffer DU. African swine fever epidemiology and control. Annu Rev Anim Biosci 2020;8:221-46.
7Niemi JK. Impacts of African swine fever on pigmeat markets in Europe. Front Vet Sci 2020;7:634.
8Weaver TR, Habib N. Evaluating Losses Associated with African Swine Fever in the People's Republic of China and Neighboring Countries. Asian Development Bank; 2020. Available from: [Last acessed on 2022 Feb 4].
9Sinelnikova EM, Dvoretskova TV, Kagan ZS. Intermediate plateaux in kinetics of the reaction catalyzed by biodegradative L-threonine dehydratase from Escherichia coli. Biokhimiia 1975;40:645-51.
10Gaudreault NN, Richt JA. Subunit vaccine approaches for African swine fever virus. Vaccines (Basel) 2019;7:E56.
11Jancovich JK, Chapman D, Hansen DT, Robida MD, Loskutov A, Craciunescu F, et al. Immunization of Pigs by DNA prime and recombinant vaccinia virus boost to identify and rank African swine fever virus immunogenic and protective proteins. J Virol 2018;92:e02219-17.
12Muthusamy K, Gopinath K, Nandhini D. Computational prediction of immunodominant antigenic regions & potential protective epitopes for dengue vaccination. Indian J Med Res 2016;144:587-91.
13Farhadi T. Effectiveness assessment of protein drugs and vaccines through in silico analysis. Biomed Biotechnol Res J 2018;2:106.
14Soria-Guerra RE, Nieto-Gomez R, Govea-Alonso DO, Rosales-Mendoza S. An overview of bioinformatics tools for epitope prediction: Implications on vaccine development. J Biomed Inform 2015;53:405-14.
15Huang Y, Niu B, Gao Y, Fu L, Li W. CD-HIT suite: A web server for clustering and comparing biological sequences. Bioinformatics 2010;26:680-2.
16Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol 2011;7:539.
17Garcia-Boronat M, Diez-Rivero CM, Reinherz EL, Reche PA. PVS: A web server for protein sequence variability analysis tuned to facilitate conserved epitope discovery. Nucleic Acids Res 2008;36:W35-41.
18Litwin S, and Jores R. Shannon Information as a Measure of Amino Acid Diversity. Theoretical and Experimental Insights into Immunology. 1992;66:279–87. Available from:
19Karosiene E, Lundegaard C, Lund O, Nielsen M. NetMHCcons: a consensus method for the major histocompatibility complex class I predictions. Immunogenetics 2012;64:177-86. Available from: doi: 10.1007/s00251-011-0579-8.
20Herrera LR, Bisa EP. In silico analysis of highly conserved cytotoxic T-cell epitopes in the structural proteins of African swine fever virus. Vet World 2021;14:2625-33.
21Prestwich A, Moore S, Kotze A, Budworth L, Lawton R, Kellar I. How can smoking cessation be induced before surgery? A systematic review and meta-analysis of behavior change techniques and other intervention characteristics. Front Psychol 2017;8:915.
22Vojdani, A., & Kharrazian, D. (2020). Potential antigenic cross-reactivity between SARS-CoV-2 and human tissue with a possible link to an increase in autoimmune diseases. Clinical immunology (Orlando, Fla.), 217, 108480.
23Halder ST, Dhorajiwala TM, Samant LR. Molecular docking studies of filarial β-tubulin protein models with antifilarial phytochemicals. Biomed Biotechnol Res J 2019;3:162-70.
24Srinivasan S, Selvaraj GF, Gopalan V, Padmanabhan P, Ramesh K, Govindan K, et al. Epitope identification and designing a potent multi-epitope vaccine construct against SARS-CoV-2 including the emerging variants. J Glob Infect Dis 2022;14:24-30.
25Lee H, Heo L, Lee MS, Seok C. GalaxyPepDock: A protein-peptide docking tool based on interaction similarity and energy optimization. Nucleic Acids Res 2015;43:W431-5.
26Yahaya AA, Sanusi S, Malo FU. Computer-assisted multi-epitopes t-cell subunit COVID-19 vaccine design. Biomed Biotechnol Res J 2021;5:27-34.
27Wang J, Youkharibache P, Zhang D, Lanczycki CJ, Geer RC, Madej T, et al. iCn3D, a web-based 3D viewer for the visualization of biomolecular structure and sequence annotation. Bioinformatics. 2020;36:131-5. Available from: doi: 10.1093/bioinformatics/btz502.
28Satpathy R, Dash N. In silico screening and characterization of novel natural peptides as spike protein inhibitors of novel coronavirus (severe acute respiratory syndrome coronavirus 2). J Prev Diagn Treat Strat Med 2022;1:54-62.
29Xue LC, Rodrigues JP, Kastritis PL, Bonvin AM, Vangone A. PRODIGY: A web server for predicting the binding affinity of protein-protein complexes. Bioinformatics 2016;32:3676-8.
30Sarkar B, Ullah MA, Araf Y, Islam NN, Zohora US. Immunoinformatics-guided designing and in silico analysis of epitope-based polyvalent vaccines against multiple strains of human coronavirus (HCoV). Expert review of vaccines 2021;1:1-21. Available from: doi: 10.1080/14760584.2021.1874925.
31Orozco M, Orellana L, Hospital A, Naganathan AN, Emperador A, Carrillo O, et al. Coarse-grained representation of protein flexibility. Foundations, successes, and shortcomings. Adv Protein Chem Struct Biol 2011;85:183-215.
32Lisowska E. The role of glycosylation in protein antigenic properties. Cell Mol Life Sci 2002;59:445-55.
33Le MT, Choi H, Lee H, Le VCQ, Ahn B, Ho CS, et al. SLA-1 genetic diversity in pigs: Extensive analysis of copy number variation, heterozygosity, expression, and breed specificity. Sci Rep 2020;10:743.
34Ando A, Kawata H, Shigenari A, Anzai T, Ota M, Katsuyama Y, et al. Genetic polymorphism of the swine major histocompatibility complex (SLA) class I genes, SLA-1, -2 and -3. Immunogenetics 2003;55:583-93.
35Hileman RE, Silvanovich A, Goodman RE, Rice EA, Holleschak G, Astwood JD, et al. Bioinformatic methods for allergenicity assessment using a comprehensive allergen database. Int Arch Allergy Immunol 2002;128:280-91.
36Aalberse RC. Autologous dust RAST. A neglected tool to detect idiopathic sources of allergens in the home. Clin Rev Allergy Immunol 2000;18:301-9.
37Herrera LR. Immunoinformatics approach in designing SARS-CoV-2 vaccine from experimentally determined SARS-CoV T-cell epitopes. J Appl Pharm Sci 2021;11:29-36.
38Chang ST, Ghosh D, Kirschner DE, Linderman JJ. Peptide length-based prediction of peptide-MHC class II binding. Bioinformatics 2006;22:2761-7.