Vaccine design is a principal application of the recently emerged biomedical field of immunoinformatics, also known as computational immunology. Algorithms may be used to model immunological data in order to map potential B- and T-lymphocyte epitopes, or antigenic determinants, of foreign proteins which elicit specific responses from an immunized individual. Determination of antibody residues is a key factor inidentifying antigen resistance mutations, designing epitope­specific probes for antibody isolation, and for developing epitope­based vaccines. Optimally, epitopes are determined at the atomic level structure of an antibody­antigen complex, although this may not in all instances be possible. Epitopes are of particular interest to both clinical and basic biomedical researchers since they hold huge potential for vaccine design, disease prevention, diagnosis and treatment. Since human immunodeficiency virus (HIV)­1 was identification in 1983, it has attracted a research impetus unrivalled in breadth and complexity, the ultimate objectives of which - a preventive vaccine and a curative treatment – remain elusive. Development of an efficacious and cost-effective vaccine is part of a strategy that is required urgently to combat acquired immunodeficiency syndrome, AIDS, the spectrum of conditions which is caused by infection with HIV-1. The challenge of predicting a viral phenotype from sequence data has considerable potential in HIV-1 vaccine design. Here, the major approaches which may be used for epitope prediction in HIV/AIDS research are discussed.

Keywords: HIV, virus, immunity, vaccine, t-cell, epitope, bioinformatics

AIDS, acquired immunodeficiency syndrome; B-cell, B lymphocyte; CTL, cytotoxic t-lymphocyte; ELISA, enzyme-linked immunosorbent assay; Env, envelope; ER, endoplasmic reticulum; HIV, human immunodeficiency virus; MHC, major histocompatibility complex; NMR, nuclear magnetic resonance; TAP, transporter associated with antigen processing; T-cell, T lymphocyte

Since immunoinformatics is still in its infancy as a scientific discipline,¹ the primary focus of this article on current prediction methods available for determining T-cell epitopes of HIV-1 antigens is to provide a primer to clinical and basic biomedical researchers who may be less familiar with these computational biology-based approaches as a powerful tool to investigate peptide vaccines.²

Antigenicity or epitope prediction approaches can be classified into one of two categories:
1. Predictors of individual TAP-binding peptides, proteasomal cleavage sites, MHC-binding regions or B- andT-cell epitopes;
2. Predictors ofseveral steps of this multi-stage process of epitope identification.

Epitope prediction complements laboratory bench experimentation and has the potential to enhance greatly research productivity by unravelling complexity, saving time and reducing financial considerations.

Determination of epitopes targeted by antibodies is valuable for understanding virus escape,³ facilitating antibody optimization^4,5 and informing epitope-based design of vaccines.⁶ Identifying the structure of antibody-antigen complexes by, for example, X-ray crystallography can provide information on an epitope at the atomic level,⁷ but such three dimensional intricacy can be technically demanding to reveal. Additional laboratory procedures for epitope delineation are available but they characterize with reduced accuracy and usually necessitate substantial experimental effort.^7‒9 Computational methods for epitope prediction originally aimed at recognizing antigenic determinants that may be recognized by any antibody epitope, and are thus not specific to a given antibody.^10‒13 More recently, immunoinformatics for antibody-specific epitope prediction (the prediction of the epitope targeted by an antibody of interest) has been developed.^7,14‒17 In particular, new protocols combine antibody-antigen neutralization data with antigen sequence information with the purpose of predicting peptide sequences that may comprise part of an epitope which stimulates an antibody of interest to vaccinologists; this is usually because its production correlates with a putatively protective immune response to a pathogen.^7,14,15

HIV/AIDS is global health concern which, through suppression of the human CD4+ T-cell immune response, renders an individual more susceptible to co-infection with tuberculosis, malaria and a number of opportunistic pathogens. In 2013, the most recent year for which statistics are available, an estimated 35million people worldwide were living with HIV, 2.1million persons were newly infected with HIV and there were around 1.5 million AIDS-related deaths.¹⁸ Cell-mediated immunity is important to human host defences against HIV-1 infection. In particular, studies of immune responses of HIV-infected patients indicate that CD8+ T-cells perform a crucial role in controlling virus load. In this regard, the onset of HIV-specific CD8+ CTL activity coincides with clearance of viraemia during primary HIV-1 infection,^19,20 and a reduced CTL response correlates with disease progression in infected individuals.^21,22 On this basis, potentiation of a putatively protective CTL response forms the basis for rational vaccine design. However, the development of a vaccine that boosts the elicitation of a broad and effective anti-viral neutralization response remains a challenge after three decades of investigation.^23­30 One of the drawbacks of traditional, inactivated vaccine candidates is a lack of broad cross-protective B- and T-cell-mediated responsiveness to HIV-1. The efficacy of those tested has been limited due to the genetic variation of the virus.³¹

Vaccine design in the pre-omics eradependedheavily on biochemical and immunological techniques such as phage display library, overlapping peptides, ELISA, immunohistochemistry, immunofluorescence, radioimmunoassay, Western blotting, NMR spectroscopy, X-ray crystallography of antibody-antigen structure, and attenuation of wild type pathogens by random mutation and serial passage, a process that is very expensive, time-consuming, with poor immunogenicity and the possibility of reversion of attenuation. Now, with the aid of epitope predictive software and databases, it is possible to narrow the focus more rapidly on a protein of interest, and thereby reduce drastically the requirement for laboratory experimentation. Hence, vaccine construction is made both cheaper and faster.¹ Given the potential importance of epitope identification in producing vaccines against infectious, immune and other antigen-related diseases, peptide sequences are studied by researchers in a wide range of biomedical fields, and as a consequence there has been a large expansion of databases, predictive methods and related software.

Structure determination, by X­ray crystallography or NMR spectroscopy, may provide atomic level resolution of epitopes and interactions in antibody­antigen complexes, but such data can be difficult or even infeasible to obtain.⁷ Cryo­electron microscopy may also be used to identify general epitope regions, but this method is frequently associated with lower resolution structures and in most instances cannot provide atomic level information.³² Epitope residue mapping may also be subjected to a range of other experimental techniques, although these are typically laborious and are restricted by various factors including sensitivity to effects from distal residues not part of the direct antibody­antigen interactions (e.g. alanine scanning) or dependence on the presence of substantial antibody interactions in a sequentially continuous region of the antigen (e.g. peptide scanning).⁸ The latter example is particularly limiting since most antibody epitopes are discontinuous, i.e. involve multiple sequential non-contiguous regions.⁷

In silico methods for epitope prediction are also available but the majority focus on predicting protein residues that form part of any epitope and thus do not recognize antibody in particular.^10‒13 Only a limited number of antibody­specific epitope prediction protocols have been proposed thus far.^16,17

The maintenance of HIV in tissue culture in the laboratory is a difficult, skilled, time-consuming process; thus, several immunoinformatics tools have been developed to predict B- and T-cell epitopes of virulence-associated proteins. These programs have emerged on the basis of accessible and validated data that have specific algorithms.^33,34 HIV/AIDS is a major public health problem for which there is a pressing need for sensitive and specific diagnostic methods. In current practice pathologists perform immunoassays as a diagnostic measure that requires purified antigen which is attained only by culture of HIV or recombinant technology. In order to circumvent this issue, bioinformatics is an alternative approach for prediction of epitopes without wet laboratory practice and which can help to accelerate virology research. For instance, T-cell epitopes may be synthesized chemically to use as antigen incommercial diagnostic kits for HIV detection and also for development of peptide-based vaccines.³⁵

The HIV Molecular Immunology Database that is maintained at the Los Alamos National Laboratory, New Mexico, USA, has collected data on HIV-specific cellular immune responses for several years and the list of targeted regions within HIV peptide sequences has grown steadily.³⁶

Investigators may choose between methods, servers and software for prediction of general epitopes, T-cell epitopes, secondary and tertiary structures, and characterization of epitopes. Currently available tools include: QuickAlign which aligns amino acids or nucleotides against known sequences;PeptGen generates overlapping peptides for any protein; PepMap generates peptide maps in Fasta, HTML and PDF formats; Hepitope scans for prospective epitopes based on HLA enrichment;HLA frequency analysis tools calculate HLA frequencies or HLA linkage disequilibrium in a population; ELF epitope location finder; Motif Scan screens alignments for HLA-binding motifs; HLA genotype/serotype, genotype/motif and supertype dictionaries; HIV/SIV Sequence Locator tool,formerly the HXB2 Numbering Engine;HIV BLAST searchespeptides against annotated HIV sequences; Mosaic Vaccine Tool Suite assesses polyvalent protein sequences for T‐cell vaccines; N‐Glycosite finds N‐linked glycosylation sites; Highlighter detects matches and mismatches in a set of aligned sequences; Genome Browser displays the HIV genome and proteome.³⁷

Analogous approaches have been utilizedpreviously to predict accurately T-cell epitopes for Mycobacterium tuberculosis,³⁸ influenza A virus,³⁹ tick-borne encephalitis virus,⁴⁰ and human papillomavirus.⁴¹ A key element of adaptive immunity is the interaction between foreign antigens and antibodies produced by B­cells. The ability to identify and characterize epitopes on antigen surfaces is important for vaccine design, development of antibody therapeutics and immunodiagnostic tests. In the last decade, significant effort has been invested to understand the nature and characteristics of linear epitopes with the goal of developing reliable methods for their prediction. Many bioinformatics analysis tools of varying utility were produced and have been reviewed.⁴²

One of the major hurdles to achieving these targets is the extraordinarycapacity of the virus for genetic variation, with some genes differing at up to 40% of nucleotide positions among tested isolates. Due to this complication, the genetic foundationfor many viral phenotypes, most prominentlysusceptibility to neutralization by a particular antibody, are troublesome to establishby computational methods.⁴³Thus, it is desirable for researchers in this field to apply these tools to their maximum capacity and to use the information they provide judiciously. Only through so doing may it be possible to preparea ‘new age’ vaccine againsta virus such as HIV-1 that undergoes constant antigenic variation.

The generation of a peptide from its precursor polypeptide is a prerequisite for the induction of a CTL response. The major cytosolic proteases associated with generating antigenic peptides, in particular the C-terminal end,reside within proteasomes.^44‒48 After proteasomal cleavage the peptide may be trimmed at the N-terminal end by other peptidases in the cytosol.⁴⁹ The next step is the translocation of the peptide from the cytosol to the interior of the ER, which is facilitated by binding to TAP. Once inside the ER, further N-terminal trimming of the peptide may occur,^47,50 as well as binding to MHC class I. The complex thus formed is transported to the cell surface, where it may be recognized by CTL.

The most restrictive step involved in antigen presentation is binding to MHC class I. It is estimated that only 1 in 200 peptides binds well enough to a given MHC class I molecule to elicit a CTL response.⁵¹ However, it has been shown that proteasomal cleavage and TAP transport efficiency both show a degree of specificity.^52,53 Reliable predictions of immunogenic peptides can reduce significantly the experimental chorein identifying new epitopes. Hence, effort has been invested to predict the outcome of the antigen presentation pathway. Several protocols have been developed that predict with great reliability the binding affinity of peptides to different MHC class I molecules.^54‒57 Likewise, programs have been designed that predict the efficiency with which peptides of arbitrary length are transported by TAP.^58,59 Assorted algorithms have also been created that aim to predictthe proteasomal cleavage pattern of proteins.^52,60‒62 In one report, an integrative approach was used to enhance CTL epitope identification.⁶³

Various novel methods have been developed by researchers to predict epitope residues and to determine antibody specificity. Broadly neutralizing antibodies against, for example, HIV­1 Envglycoprotein,^27,64‒66 can have utility as therapeutics in the context of passive transfer⁶⁷ and as templates for the design of epitope-specific vaccines.⁶⁸ The identification of an epitope targeted by an antibody can aid understanding of virus resistance and escape mutations,³ provide pointers to improve antibody affinity,^4,5 and direct immunogen design to focus better the immune response on virus-neutralizing epitopes.⁶

Computational docking may be used to predict epitope residues by means of generating a structural model of the antibody­antigen complex. However, docking depends on the existence of separate,accurate structural models of antibody and antigen, and scoring functions are, in general, suboptimal^69,70 and in most cases a poor predictor of the epitope of interest.⁷¹ Recently, a computational method was proposed for predicting the epitopes of putative HIV-1 antigens based on the similarity of their neutralization fingerprints to those of antigens with known epitopes.⁷² This technique, however, does not provide residue level information and is not applicable to antibodies that bind to novel epitopes. Another contemporary study utilized HIV­1 antibody neutralization panels to identify antigen residues that are functionally important for binding of specific antibodies.¹⁴ In another study, researchers described an efficient computational method to predict antibody­specific HIV­1 Env epitopes at the residue level based on neutralization panels of diverse viral strains.⁹ This algorithm is intrinsically relevant to antigens that demonstrate sequence diversity and its accuracy was shown to correlate inversely with sequence conservation of the examined epitope.

In further significant progress, researchers integrated global sequence and immunology databases to investigate systematically the connection between HIV-1 amino acid sequences and CTL epitope distributions.⁹ CTL responses to five HIV-1 proteins - Env, Nef, Gag p17, Gag p24 and reverse transcriptase - have been characterized especially well in the published literature. By comparing the distribution of CTL epitopes in these proteins to global protein sequence alignments (such as those retrieved from NCBI GenBank), distinct characteristics of HIV amino acid sequences that correlate with CTL epitope localization were identified. HIV CTL epitopes defined experimentally are concentrated in relatively conserved regions. Highly variable regions that lack epitopes show cumulative evidence of prior immune escape that may render them relatively refractive to CTL reactivity through a scarcity of predicted proteasome-processing sites and an accumulation of amino acids that do not serve as C-terminal anchor residues.⁷³ Furthermore, CTL epitopes are frequently concentrated in alpha-helical regions of proteins. Based on amino acid sequence characteristics, regions in HIV regulatory and accessory proteins were identified that would be likely to contain CTL epitopes; these predictions were subsequently validated by comparison to new sets of experimentally defined epitopes of HIV-1.⁷³

The design of protective immunogens presents a significant challenge because of the great ability of viruses to evade host immunity.^74­76 Ongoing HIV­1 vaccine research includes determining and characterizing broadly neutralizing antibodies and the epitopes they target.^77,78 Hence, identifying antigenic targets of such antibodies, in parallel with mapping the immunologically important residues of known epitopes that affect neutralization, is a major objective. Env is a highly variable protein and consequently identification of key residues that affect neutralization is not a straightforward task. A lack of neutralization may sometimes be explained by amino acid changes in known epitopes, but often epitope conservation does not ensure neutralization.⁷⁹ Many regions external to known epitopes have been shown to affect neutralization sensitivity.⁸⁰ Identification of B- and T-cell epitopes is a crucial step in peptide vaccine design but which does not enjoy sufficient accuracy due to the extent of mutations in HIV-1 observed in regionally distinct isolates. Thus, prediction of epitope structures with a wider range of variationmay facilitatemore efficient treatment of HIV/AIDS on a global scale.

A key consideration when designing an HIV/AIDS vaccine is how its structure may best promote elicitation of a protective immune response. This includes identification of novel CTL epitopes with a high binding affinity for MHC class I. Utilization of bioinformatics for accurate prediction of epitope identity and characteristics enables a next generation vaccine design to incorporate appropriate residues to stimulate antigen-specific B- and/or T-cells. As well as for the development of a vaccine an urgent need exists to establish sensitive antigen-based immunodiagnostics for earlier monitoring of HIV. Ongoing and prospective research should endeavour to design a candidate vaccine structure to include several variation spredicated on information combined from computational biology and reliable traditional techniques such as X-ray crystallography. Sequence data required may be retrieved from current genome and protein databases to predict a limited number of functionally important antigen residues of an examined antibody, rather than to identify the antibody epitope per se.

The incorporation of data from geographically diverse strains of the same virus may increase the likelihood of observing an association between changes in neutralization potency and variations in antigenic sequence.This suggests that the combination of a neutralization­based method with an immunoinformatics approach that provides structural information could prove a useful strategy for the accurate prediction of antibody epitopes of genetically diverse HIV antigens.

Immunoinformatics is one of a new wave of computational approaches which are used in identification of antigenic peptides and which can be applied to development of diagnostics, therapeutics and vaccines. In recent years the CTL response to HIV-1 has been studied intensively and hundreds of T-cell epitopes are now defined experimentally, their sequences published and compiled in the HIV Molecular Immunology Database. An increasing volume and complexity of information about conserved epitopes of HIV-1 brings closer the development of a universal epitope­based vaccine. Such a construct should contain both conserved B-cell epitopes that are important for induction of cross-protective antibodies and CTL epitopes for involvement of the cellular immune response. Future research should aim to utilise immunoinformatics as a tool to define characteristics of HIV-1 antigens, to identify T-cell epitopes located at amino acid residues and to evaluate the antigenic capacity of predicted T-cell epitope constructs. The novel insights provided from such bioinformatics application would facilitate the production of locally or globally effective antigen-based diagnostics and peptide-based vaccines designed to combat HIV/AIDS.

None.

Author declares there are no conflicts of interest.

None.

1. De Groot AS. Immunomics: discovering new targets for vaccines and therapeutics. Drug Discov Today. 2006;11(5‒6):203‒209.
2. Mathew S, Qadri I. Updates on peptide vaccines. Int J Vaccines Vaccin. 2015;1(1):00001.
3. Adams O, Bonzel L, Kovacevic A, et al. Palivizumab-resistant human respiratory syncytial virus infection in infancy. Clin Infect Dis. 2010;51(2):185‒188.
4. Diskin R, Scheid JF, Marcovecchio PM, et al. Increasing the potency and breadth of an HIV antibody by using structure-based rational design. Science. 2011;334(6060):1289‒1293.
5. Yu CM, Peng HP, Chen IC, et al. Rationalization and design of the complementarity determining region sequences in an antibody-antigen recognition interface. PLoS One. 2012;7(3):e33340.
6. Ofek G, Guenaga FJ, Schief WR, et al. Elicitation of structure-specific antibodies by epitope scaffolds. Proc Natl Acad Sci USA. 2010;107(42):17880‒17887.
7. Gershoni JM, Roitburd-Berman A, Siman-Tov DD, et al. Epitope mapping: the first step in developing epitope-based vaccines. BioDrugs. 2007;21(3):145‒156.
8. Geysen HM, Meloen RH, Barteling SJ. Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc Natl Acad Sci USA. 1984;81(13):3998‒4002.
9. Chuang G-Y, Acharya P, Schmidt SD, et al. Residue-level prediction of HIV-1 antibody epitopes based on neutralization of diverse viral strains. J Virol. 2013;87(18):10047‒10058.
10. Parker J, Guo D, Hodges RS. New hydrophilicity scale derived from high-performance liquid chromatography peptide retention data: correlation of predicted surface residues with antigenicity and X-ray-derived accessible sites. Biochemistry. 1986;25(19):5425‒5432.
11. Kulkarni-Kale U, Bhosle S, Kolaskar AS. CEP: a conformational epitope prediction server. Nucleic Acids Res. 2005;33:W168‒W171.
12. Ponomarenko J, Bui HH, Li W, et al. ElliPro: a new structure-based tool for the prediction of antibody epitopes. BMC Bioinformatics. 2008;9:514.
13. Haste Andersen P, Nielsen M, Lund O. Prediction of residues in discontinuous B‐cell epitopes using protein 3D structures. Protein Sci. 2006;15(11):2558‒2567.
14. West AP Jr, Scharf L, Horwitz J, et al. Computational analysis of anti-HIV-1 antibody neutralization panel data to identify potential functional epitope residues. Proc Natl Acad Sci USA. 2013;110(26):10598‒10603.
15. Ferguson AL, Falkowska E, Walker LM, et al. Computational prediction of broadly neutralizing HIV-1 antibody epitopes from neutralization activity data. PLoS One. 2013;8(12):e80562.
16. Zhao L, Wong L, Li J. Antibody­specified B­cell epitope prediction in line with the principle of context­awareness. IEEE/ACM Trans Comput Biol Bioinform. 2011;8(6):1483‒1494.
17. Ahmad S, Mizuguchi K. Partner-aware prediction of interacting residues in protein-protein complexes from sequence data. PLoS One. 2011;6(12):e29104.
18. Granich R, Gupta S, Hersh B, et al. Trends in AIDS Deaths, New Infections and ART Coverage in the Top 30 Countries with the Highest AIDS Mortality Burden; 1990-2013. PLoS One. 2015;10(7):e0131353.
19. Borrow P, Lewicki H, Hahn BH, et al. Virus-specific CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection. J Virol. 1994;68(9):6103‒6110.
20. Koup RA, Safrit JT, Cao Y, et al. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J Virol. 1994;68(7):4650‒4655.
21. Carmichael A, Jin X, Sissons P, et al. Quantitative analysis of the human immunodeficiency virus type 1 (HIV-1)-specific cytotoxic T lymphocyte (CTL) response at different stages of HIV-1 infection: differential CTL responses to HIV-1 and Epstein-Barr virus in late disease. J Exp Med. 1993;177(2):249‒256.
22. Klein MR, van Baalen CA, Holwerda AM, et al. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J Exp Med. 1995;181(7):1365‒1372.
23. Burton DR, Stanfield RL, Wilson IA. Antibody vs. HIV in a clash of evolutionary titans. Proc Natl Acad Sci USA. 2005;102(42):14943‒14948.
24. Flynn NM, Forthal DN, Harro CD, et al. Placebo-controlled phase 3 trial of a recombinant glycoprotein 120 vaccine to prevent HIV-1 infection. J Infect Dis. 2005;191(5):654‒665.
25. Pitisuttithum P, Gilbert P, Gurwith M, et al. Randomized, double-blind, placebo-controlled efficacy trial of a bivalent recombinant glycoprotein 120 HIV-1 vaccine among injection drug users in Bangkok, Thailand. J Infect Dis. 2006;194(12):1661‒1671.
26. Saphire EO, Montero M, Menendez A, et al. Structure of a high-affinity "mimotope" peptide bound to HIV-1-neutralizing antibody b12 explains its inability to elicit gp120 cross-reactive antibodies. J Mol Biol . 2007;369(3):696‒709.
27. Walker LM, Phogat SK, Chan-Hui PY, et al. Broad and potent neutralizing antibodies from an African donor reveal a new HIV-1 vaccine target. Science. 2009;326(5950):285‒289.
28. Burton DR, Weiss RA. AIDS/HIV. A boost for HIV vaccine design. Science. 2010;329(5993):770‒773.
29. Virgin HW, Walker BD. Immunology and the elusive AIDS vaccine. Nature. 2010;464(7286):224‒231.
30. Walker LM, Huber M, Doores KJ, et al. Broad neutralization coverage of HIV by multiple highly potent antibodies. Nature. 2011;477(7365):466‒470.
31. Excler JL, Robb ML, Kim JH. Prospects for a globally effective HIV-1 vaccine. Vaccine. 2015;(15):00378‒00383.
32. Glaeser RM, Hall RJ. Reaching the information limit in cryo-EM of biological macromolecules: experimental aspects. Biophys J. 2011;100(10):2331‒2337.
33. Petrovsky N, Brusic V. Computational immunology: the coming of age. Immunol Cell Biol. 2002;80(3):248‒254.
34. Petrovsky N, Silva D, Brusic V. The future for computational modelling and prediction systems in clinical immunology. Novartis Found Symp. 2003;254: 23‒32.
35. Chaturvedi D, Pandey M, Singh V, et al. Immunoinformatics approach to T-cell epitopes in human immunodeficiency virus. J Environ Sci Technol. 2013;1(3):57‒61.
36. Addo MM, Altfeld M, Rosenberg ES, et al. The HIV-1 regulatory proteins Tat and Rev are frequently targeted by cytotoxic T lymphocytes derived from HIV-1-infected individuals. Proc Natl Acad Sci USA. 2001;98(4):1781‒1786.
37. HIV Molecular Immunology Database.
38. Somvanshi P, Singh V, Seth PK. In silico prediction of epitopes in virulence proteins of Mycobacterium tuberculosis H37Rv for diagnostic and subunit vaccine design. J Proteomics Bioinform. 2008;1:143‒153.
39. Somvanshi P, Singh V, Seth PK. Prediction of epitopes in hemagglutinin and neuraminidase proteins of influenza A virus H5N1 strain: a clue for diagnostic and vaccine development. OMICS. 2008;12(1):61‒69.
40. Chaudhary DK, Mani I, Singh V. Identification of T-cell epitopes in structural proteins of Tick borne encephalitis virus for vaccine development. Appl Bio Pharma Tech. 2012;3(2):39‒43.
41. Chaudhary DK, Mani I, Singh V. Identification of T-cell antigenic determinants in the proteins of Human Papillomavirus. Adv Biotech Bioinform. 2012;1(1):79‒86.
42. Evans MC. Recent advances in immunoinformatics: application of in silico tools to drug development. Curr Opin Drug Discov Devel. 2008;11(2):233‒241.
43. Hepler NL, Scheffler K, Weaver S, et al. IDEPI: rapid prediction of HIV-1 antibody epitopes and other phenotypic features from sequence data using a flexible machine learning platform. PLoS Comput Biol. 2014;10(9):e1003842.
44. Craiu A, Akopian T, Goldberg A, et al. Two distinct proteolytic processes in the generation of a major histocompatibility complex class I-presented peptide. Proc Natl Acad Sci USA . 1997;94(20):10850‒10855.
45. Stoltze L, Dick TP, Deeg M, et al. Generation of the vesicular stomatitis virus nucleoprotein cytotoxic T lymphocyte epitope requires proteasome‐dependent and‐independent proteolytic activities. Eur J Immunol. 1998;28(12):4029‒4036.
46. Paz P, Brouwenstijn N, Perry R, et al. Discrete proteolytic intermediates in the MHC class I antigen processing pathway and MHC I-dependent peptide trimming in the ER. Immunity. 1999;11(2):241‒251.
47. Mo X, Cascio P, Lemerise K, et al. Distinct proteolytic processes generate the C and N termini of MHC class I-binding peptides. J Immunol. 1999;163(11):5851‒5859.
48. Altuvia Y, Margalit H. Sequence signals for generation of antigenic peptides by the proteasome: implications for proteasomal cleavage mechanism. J Mol Biol. 2000; 295(4):879‒890.
49. Lévy F, Burri L, Morel S, et al. The final N-terminal trimming of a subaminoterminal proline-containing HLA class I-restricted antigenic peptide in the cytosol is mediated by two peptidases. J Immunol. 2002;169(8):4161‒4171.
50. Stoltze L, Schirle M, Schwarz G, et al. Two new proteases in the MHC class I processing pathway. Nat Immunol. 2000;1(5):413‒418.
51. Yewdell JW, Bennink JR. Immunodominance in major histocompatibility complex class I-restricted T lymphocyte responses. AnnuRev Immunol. 1999;17:51‒88.
52. Keşmir C, Nussbaum AK, Schild H, et al. Prediction of proteasome cleavage motifs by neural networks. Protein Eng. 2002;15(4):287‒296.
53. van Endert PM, Riganelli D, Greco G, Fleischhauer K, Sidney J, et al. (1995) The peptide-binding motif for the human transporter associated with antigen processing. J Exp Med 182(6): 1883-1895.
54. Brusic V, Rudy G, Harrison LC. Prediction of MHC binding peptides using artificial neural networks. In: Stonier RJ, Yu XS (Eds.), Complex Systems: Mechanism of Adaptation. IOS Press, Amsterdam, Netherlands. 1994. p.253‒260.
55. Buus S, Lauemøller SL, Worning P, et al. Sensitive quantitative predictions of peptide‐MHC binding by a ‘Query by Committee’artificial neural network approach. Tissue Antigens. 2003;62(5):378‒384.
56. Nielsen M, Lundegaard C, Worning P, et al. Reliable prediction of T‐cell epitopes using neural networks with novel sequence representations. Protein Sci. 2003;12(5):1007‒1017.
57. Nielsen M, Lundegaard C, Worning P, et al.  Bioinformatics. 2004;20(9):1388‒1397.
58. BrusicV, van Endert P, Zeleznikow J, et al. A neural network model approach to the study of human TAP transporter. In Silico Biol . 1998;1(2):109‒121.
59. Peters B, Bulik S, Tampe R, et al. Identifying MHC class I epitopes by predicting the TAP transport efficiency of epitope precursors. J Immunol. 2003;171(4):1741‒1749.
60. Nielsen M, Lundegaard C, Lund O, et al. The role of the proteasome in generating cytotoxic T-cell epitopes: insights obtained from improved predictions of proteasomal cleavage. Immunogenetics. 2005;57(1‒2):33‒41.
61. Sette A, Sidney J. HLA supertypes and supermotifs: a functional perspective on HLA polymorphism. Curr Opin Immunol. 1998;10(4):478‒482.
62. Lund O, Nielsen M, Kesmir C, et al. Definition of supertypes for HLA molecules using clustering of specificity matrices. Immunogenetics. 2004;55(12):797‒810.
63. Larsen MV, Lundegaard C, Lamberth K, et al. An integrative approach to CTL epitope prediction: a combined algorithm integrating MHC class I binding, TAP transport efficiency, and proteasomal cleavage predictions. Eur J Immunol . 2005;35(8):2295‒2303.
64. Corti D, Langedijk JP, Hinz A, et al. Analysis of memory B cell responses and isolation of novel monoclonal antibodies with neutralizing breadth from HIV-1-infected individuals. PLoS One. 2010;5(1):e8805.
65. Scheid JF, Mouquet H, Ueberheide B, et al. Sequence and structural convergence of broad and potent HIV antibodies that mimic CD4 binding. Science. 2011;333(6049):1633‒1637.
66. Wu X, Yang ZY, Li Y, et al. Rational design of envelope identifies broadly neutralizing human monoclonal antibodies to HIV-1. Science. 2010;329(5993):856‒861.
67. Klein F, Halper-Stromberg A, Horwitz JA, et al. HIV therapy by a combination of broadly neutralizing antibodies in humanized mice. Nature. 2012;492(7427):118‒122.
68. Sette A, Fikes J. Epitope­based vaccines: an update on epitope identification, vaccine design and delivery. Curr Opin Immunol. 2003;15(4):461‒470.
69. Halperin I, Ma B, Wolfson H, et al. Principles of docking: an overview of search algorithms and a guide to scoring functions. Proteins. 2002;47(4):409‒443.
70. Vajda S, Camacho CJ. Protein­protein docking: is the glass half­full or half­empty? Trends Biotechnol. 2004;22(3):110‒116.
71. Wodak SJ, Méndez R. Prediction of protein-protein interactions: the CAPRI experiment, its evaluation and implications. Curr Opin Struct Biol. 2004;14(2):242‒249.
72. Georgiev IS, Doria-Rose NA, Zhou T, et al. Delineating antibody recognition in polyclonal sera from patterns of HIV-1 isolate neutralization. Science. 2013;340(6133): 751‒756.
73. Yusim K, Kesmir C, Gaschen B, et al. Clustering patterns of cytotoxic T-lymphocyte epitopes in human immunodeficiency virus type 1 (HIV-1) proteins reveal imprints of immune evasion on HIV-1 global variation. J Virol. 2002;76(17):8757‒8768.
74. Richman DD, Wrin T, Little SJ, et al. Rapid evolution of the neutralizing antibody response to HIV type 1 infection. Proc Natl Acad Sci USA. 2003;100(7):4144‒4149.
75. Wei X, Decker JM, Wang S, et al. Antibody neutralization and escape by HIV-1. Nature. 2003;422(6929):307‒312.
76. Morgand M, Morgand M, Moreau A, et al. Evidence for a continuous drift of the HIV­1 species towards higher resistance to neutralizing antibodies over the course of the epidemic. PLoS Pathog. 2013;9(7):e1003477.
77. Jardine J, Julien JP, Menis S, et al. Rational HIV immunogen design to target specific germline B cell receptors. Science. 2013;340(6133):711‒716.
78. Van Regenmortel MH. Basic research in HIV vaccinology is hampered by reductionist thinking. Front Immunol. 2012;3:194.
79. Binley JM, Wrin T, Korber B, et al. Comprehensive cross-clade neutralization analysis of a panel of anti-human immunodeficiency virus type 1 monoclonal antibodies. J Virol. 2004;78(23):13232‒13252.
80. Pantophlet R, Ollmann Saphire E, Poignard P, et al. Fine mapping of the interaction of neutralizing and nonneutralizing monoclonal antibodies with the CD4 binding site of human immunodeficiency virus type 1 gp120. J Virol. 2003;77(1):642‒658.