Benchmarking B cell epitope prediction: underperformance of existing methods

The self/nonself issue: A confrontation between proteomes

Defining self and nonself is the most compelling challenge in science today, at the basis of the numerous questions that remain unanswered in the immunology-pathology-therapy debate. The generation of the antibody repertoire, the complicated scenario offered by tolerance and autoimmunity, natural auto-antibodies and their relationship to autoimmune diseases, and positive and negative selection are only a few examples of the unresolved immunological questions. In this context, we proposed that sequence similarity to the host proteome modulates antigen peptide recognition and immunogenicity. Using the available proteome assemblies of viruses, bacteria and higher vertebrates, and applying the low-similarity criterion, we are systematically defining the proteomic similarity of B-cell epitopes already validated experimentally. Here, we report further data documenting that a low similarity to the host proteome is the common property that defines the immunological "nonself" nature of antigenic sequences in cancer, autoimmunity, infectious diseases and allergy.

Prediction of antibody response using recombinant human protein fragments as antigen

A great need exists for prediction of antibody response for the generation of antibodies toward protein targets. Earlier studies have suggested that prediction methods based on hydrophilicity propensity scale, in which the degree of exposure of the amino acid in an aqueous solvent is calculated, has limited value. Here, we show a comparative analysis based on 12,634 affinity-purified antibodies generated in a standardized manner against human recombinant protein fragments. The antibody response (yield) was measured and compared to theoretical predictions based on a large number (544) of published propensity scales. The results show that some of the scales have predictive power, although the overall Pearson correlation coefficient is relatively low (0.2) even for the best performing amino acid indices. Based on the current data set, a new propensity scale was calculated with a Pearson correlation coefficient of 0.25. The values correlated in some extent to earlier scales, including large penalty for hydrophobic and cysteine residues and high positive contribution from acidic residues, but with relatively low positive contribution from basic residues. The fraction of immunogens generating low antibody responses was reduced from 30% to around 10% if immunogens with a high propensity score (>0.48) were selected as compared to immunogens with lower scores (<0.29). The study demonstrates that a propensity scale might be useful for prediction of antibody response generated by immunization of recombinant protein fragments. The data set presented here can be used for further studies to design new prediction tools for the generation of antibodies to specific protein targets.

Peptide binding to the HLA-DRB1 supertype: a proteochemometrics analysis

A proteochemometrics approach was applied to a set of 2666 peptides binding to 12 HLA-DRB1 proteins. Sequences of both peptide and protein were described using three z-descriptors. Cross terms accounting for adjacent positions and for every second position in the peptides were included in the models, as well as cross terms for peptide/protein interactions. Models were derived based on combinations of different blocks of variables. These models had moderate goodness of fit, as expressed by r2, which ranged from 0.685 to 0.732; and good cross-validated predictive ability, as expressed by q2, which varied from 0.678 to 0.719. The external predictive ability was tested using a set of 356 HLA-DRB1 binders, which showed an r2(pred) in the range 0.364-0.530. Peptide and protein positions involved in the interactions were analyzed in terms of hydrophobicity, steric bulk and polarity.

Removal of B cell epitopes as a practical approach for reducing the immunogenicity of foreign protein-based therapeutics

Immunogenicity of non-human proteins with useful therapeutic properties has prevented their development for use in the therapy of disease. However, this class of proteins could be very useful, if their immunogenicity could be markedly reduced so that many treatment cycles could be administered. One approach to reduce the immunogenicity of foreign proteins is to identify B cell epitopes on the protein and eliminate them by mutagenesis. In this article, theoretical aspects and experimental evidence for the feasibility of B cell epitope removal is reviewed. A special focus is given to our results with deimmunization of recombinant immunotoxins in which Fvs are fused to a 38kDa portion of the bacterial protein, Pseudomonas exotoxin A (PE38). Immunotoxins targeting CD22 and CD25 have produced complete remissions in many patients with drug resistant Hairy Cell Leukemia and are being evaluated in other malignancies. Experimental data summarized in this review indicates that removal of B cell epitopes is a practical approach for making less immunogenic protein therapeutics from non-human functional proteins. This approach requires grouping of the epitopes to identify targets for deimmunization followed by quantitative analysis of the decrease in affinity produced by the mutations in B cell epitopes.

Prediction of antigenic epitopes on protein surfaces by consensus scoring

Background

Prediction of antigenic epitopes on protein surfaces is important for vaccine design. Most existing epitope prediction methods focus on protein sequences to predict continuous epitopes linear in sequence. Only a few structure-based epitope prediction algorithms are available and they have not yet shown satisfying performance.

Results

We present a new antigen Epitope Prediction method, which uses ConsEnsus Scoring (EPCES) from six different scoring functions - residue epitope propensity, conservation score, side-chain energy score, contact number, surface planarity score, and secondary structure composition. Applied to unbounded antigen structures from an independent test set, EPCES was able to predict antigenic eptitopes with 47.8% sensitivity, 69.5% specificity and an AUC value of 0.632. The performance of the method is statistically similar to other published methods. The AUC value of EPCES is slightly higher compared to the best results of existing algorithms by about 0.034.

Conclusion

Our work shows consensus scoring of multiple features has a better performance than any single term. The successful prediction is also due to the new score of residue epitope propensity based on atomic solvent accessibility.

Biochemical and structural characterization of alanine racemase from Bacillus anthracis (Ames)

BACKGROUND

Bacillus anthracis is the causative agent of anthrax and a potential bioterrorism threat. Here we report the biochemical and structural characterization of B. anthracis (Ames) alanine racemase (AlrBax), an essential enzyme in prokaryotes and a target for antimicrobial drug development. We also compare the native AlrBax structure to a recently reported structure of the same enzyme obtained through reductive lysine methylation.

RESULTS

B. anthracis has two open reading frames encoding for putative alanine racemases. We show that only one, dal1, is able to complement a D-alanine auxotrophic strain of E. coli. Purified Dal1, which we term AlrBax, is shown to be a dimer in solution by dynamic light scattering and has a Vmax for racemization (L- to D-alanine) of 101 U/mg. The crystal structure of unmodified AlrBax is reported here to 1.95 A resolution. Despite the overall similarity of the fold to other alanine racemases, AlrBax makes use of a chloride ion to position key active site residues for catalysis, a feature not yet observed for this enzyme in other species. Crystal contacts are more extensive in the methylated structure compared to the unmethylated structure.

CONCLUSION

The chloride ion in AlrBax is functioning effectively as a carbamylated lysine making it an integral and unique part of this structure. Despite differences in space group and crystal form, the two AlrBax structures are very similar, supporting the case that reductive methylation is a valid rescue strategy for proteins recalcitrant to crystallization, and does not, in this case, result in artifacts in the tertiary structure.

Bioinformatics for study of autoimmunity

Recent years have witnessed an explosive growth in available biological data pertaining to autoimmunity research. This includes a tremendous quantity of sequence data (biological structures, genetic and physical maps, pathways, etc.) generated by genome and proteome projects plus extensive clinical and epidemiological data. Autoimmunity research stands to greatly benefit from this data so long as appropriate strategies are available to enable full access to and utilization of this data. The quantity and complexity of this biological data necessitates use of advanced bioinformatics strategies for its efficient retrieval, analysis and interpretation. Major progress has been made in development of specialized tools for storage, analysis and modeling of immunological data, and this has led to development of a whole new field know as immunoinformatics. With advances in novel high-throughput immunology technologies immunoinformatics is transforming understanding of how the immune system functions. This paper reviews advances in the field of immunoinformatics pertinent to autoimmunity research including databases, tools in genomics and proteomics, tools for study of B- and T-cell epitopes, integrative approaches, and web servers.

Immunoinformatics: current trends and future directions

Immunoinformatics has recently emerged as a critical field for accelerating immunology research. Although still an evolving process, computational models now play instrumental roles, not only in directing the selection of key experiments, but also in the formulation of new testable hypotheses through detailed analysis of complex immunologic data that could not be achieved using traditional approaches alone. Immunomics, which combines traditional immunology with computer science, mathematics, chemistry, biochemistry, genomics and proteomics for the large-scale analysis of immune system function, offers new opportunities for future bench-to-bedside research. In this article, we review the latest trends and future directions of the field.

An in vivo protective response against toxic effects of the dermonecrotic protein from Loxosceles intermedia spider venom elicited by synthetic epitopes

Loxoscelism is a necrotic-hemolytic syndrome caused by bites of brown spiders belonging to the genus Loxosceles. Many approaches for the treatment of Loxosceles poisoning have already been proposed, among which administration of specific antivenom is thought to be the more specific. We have evaluated the use of peptides as immunogen to raise in rabbits an antibody response that could protect animals from a challenge by the Loxtox isoform LiD1, one of the main toxic component of Loxosceles intermedia venom. Six antigenic regions of LiD1 were mapped by using the SPOT method. The corresponding peptides were further chemically synthesized, mixed, and used as immunogens in rabbits. Control animal received recombinant LiD1 alone or together with peptides. We found that the rabbit antibody response to peptides was cross-reactive with LiD1, although only one peptide from the mix of six was immunogenic. The dermonecrotic, hemorrhagic and oedema forming activities induced by LiD1 in naïve rabbits were inhibited by 82%, 35% and 35% respectively, by preincubation of LiD1 with anti-peptide antibodies prepared from immunized rabbits. Animals that were immunized with peptides or LiD1r, were found to be protected from the dermonecrotic, hemorrhagic and oedema forming activities induced by a challenge with LiD1. The protection conferred by peptides was, however, lower than that provided by the peptide protein combination or by the full-length protein. These results encourage us in the utilization of synthetic peptides for therapeutic serum development or vaccination approaches.

Estimation and extraction of B-cell linear epitopes predicted by mathematical morphology approaches

B-cell epitope prediction facilitates the design and synthesis of short peptides for various immunological applications. Several algorithms have been developed to predict B-cell linear epitopes (LEs) from primary sequences of antigens, providing important information for immunobiological experiments and antibody design. This paper describes two robust methods, LE prediction with/without local peak extraction (LEP-LP and LEP-NLP), based on antigenicity scale and mathematical morphology for the prediction of B-cell LEs. Previous studies revealed that LEs could occur in regions with low-to-moderate but not globally high antigenicity scales. Hence, we developed a method adopting mathematical morphology to extract local peaks from a linear combination of the propensity scales of physico-chemical characteristics at each antigen residue. Comparison among LEP-LP/LEP-NLP, BepiPred and BEPITOPE revealed that our algorithms performed better in retrieving epitopes with low-to-moderate antigenicity and achieved comparable performance according to receiver operation characteristics (ROC) curve analysis. Of the identified LEs, over 30% were unable to be predicted by BepiPred and BEPITOPE employing an average threshold of antigenicity index or default settings. Our LEP-LP method provides a bioinformatics approach for predicting B-cell LEs with low- to-moderate antigenicity. The web-based server was established at http://biotools.cs.ntou.edu.tw/lepd_antigenicity. php for free use.

Prediction of linear B-cell epitopes

The prediction of B-cell epitopes is desirable for designing peptide-based vaccines, or generating antibodies especially if the purified protein is difficult to obtain and immunization has to be performed with protein-derived synthetic peptides. A number of freely available tools predict epitopes from protein sequence or structural information. The handling of these tools is described and the predictive power is assessed using test data based on the proteome of HIV, where comprehensive epitope mapping data are available.

What is a B-cell epitope?

The antigenicity of proteins resides in different types of antigenic determinants known as continuous and discontinuous epitopes, cryptotopes, neotopes, and mimotopes. All epitopes have fuzzy boundaries and can be identified only by their ability to bind to certain antibodies. Antigenic cross-reactivity is a common phenomenon because antibodies are always able to recognize a considerable number of related epitopes. This places severe limits to the specificity of antibodies. Antigenicity, which is the ability of an epitope to react with an antibody, must be distinguished from its immunogenicity or ability to induce antibodies in a competent vertebrate host. Failure to make this distinction partly explains why no successful peptide-based vaccines have yet been developed. Methods for predicting the epitopes of proteins are discussed and the reasons for the low success rate of epitope prediction are analyzed.

Pep-3D-Search: a method for B-cell epitope prediction based on mimotope analysis

Background

The prediction of conformational B-cell epitopes is one of the most important goals in immunoinformatics. The solution to this problem, even if approximate, would help in designing experiments to precisely map the residues of interaction between an antigen and an antibody. Consequently, this area of research has received considerable attention from immunologists, structural biologists and computational biologists. Phage-displayed random peptide libraries are powerful tools used to obtain mimotopes that are selected by binding to a given monoclonal antibody (mAb) in a similar way to the native epitope. These mimotopes can be considered as functional epitope mimics. Mimotope analysis based methods can predict not only linear but also conformational epitopes and this has been the focus of much research in recent years. Though some algorithms based on mimotope analysis have been proposed, the precise localization of the interaction site mimicked by the mimotopes is still a challenging task.

Results

In this study, we propose a method for B-cell epitope prediction based on mimotope analysis called Pep-3D-Search. Given the 3D structure of an antigen and a set of mimotopes (or a motif sequence derived from the set of mimotopes), Pep-3D-Search can be used in two modes: mimotope or motif. To evaluate the performance of Pep-3D-Search to predict epitopes from a set of mimotopes, 10 epitopes defined by crystallography were compared with the predicted results from a Pep-3D-Search: the average Matthews correlation oefficient (MCC), sensitivity and precision were 0.1758, 0.3642 and 0.6948. Compared with other available prediction algorithms, Pep-3D-Search showed comparable MCC, specificity and precision, and could provide novel, rational results. To verify the capability of Pep-3D-Search to align a motif sequence to a 3D structure for predicting epitopes, 6 test cases were used. The predictive performance of Pep-3D-Search was demonstrated to be superior to that of other similar programs. Furthermore, a set of test cases with different lengths of sequences was constructed to examine Pep-3D-Search's capability in searching sequences on a 3D structure. The experimental results demonstrated the excellent search capability of Pep-3D-Search, especially when the length of the query sequence becomes longer; the iteration numbers of Pep-3D-Search to precisely localize the target paths did not obviously increase. This means that Pep-3D-Search has the potential to quickly localize the epitope regions mimicked by longer mimotopes.

Conclusion

Our Pep-3D-Search provides a powerful approach for localizing the surface region mimicked by the mimotopes. As a publicly available tool, Pep-3D-Search can be utilized and conveniently evaluated, and it can also be used to complement other existing tools. The data sets and open source code used to obtain the results in this paper are available on-line and as supplementary material. More detailed materials may be accessed at .

COBEpro: a novel system for predicting continuous B-cell epitopes

Accurate prediction of B-cell epitopes has remained a challenging task in computational immunology despite several decades of research. Only 10% of the known B-cell epitopes are estimated to be continuous, yet they are often the targets of predictors because a solved tertiary structure is not required and they are integral to the development of peptide vaccines and engineering therapeutic proteins. In this article, we present COBEpro, a novel two-step system for predicting continuous B-cell epitopes. COBEpro is capable of assigning epitopic propensity scores to both standalone peptide fragments and residues within an antigen sequence. COBEpro first uses a support vector machine to make predictions on short peptide fragments within the query antigen sequence and then calculates an epitopic propensity score for each residue based on the fragment predictions. Secondary structure and solvent accessibility information (either predicted or exact) can be incorporated to improve performance. COBEpro achieved a cross-validated area under the curve (AUC) of the receiver operating characteristic up to 0.829 on the fragment epitopic propensity scoring task and an AUC up to 0.628 on the residue epitopic propensity scoring task. COBEpro is incorporated into the SCRATCH prediction suite at http://scratch.proteomics.ics.uci.edu.

A machine-learning approach for predicting B-cell epitopes

The immune activity of an antibody is directed against a specific region on its target antigen known as the epitope. Numerous immunodetection and immunotheraputics applications are based on the ability of antibodies to recognize epitopes. The detection of immunogenic regions is often an essential step in these applications. The experimental approaches used for detecting immunogenic regions are often laborious and resource-intensive. Thus, computational methods for the prediction of immunogenic regions alleviate this drawback by guiding the experimental procedures. In this work we developed a computational method for the prediction of immunogenic regions from either the protein three-dimensional structure or sequence when the structure is unavailable. The method implements a machine-learning algorithm that was trained to recognize immunogenic patterns based on a large benchmark dataset of validated epitopes derived from antigen structures and sequences. We compare our method to other available tools that perform the same task and show that it outperforms them.

Predicting linear B-cell epitopes using string kernels

The identification and characterization of B‐cell epitopes play an important role in vaccine design, immunodiagnostic tests, and antibody production. Therefore, computational tools for reliably predicting linear B‐cell epitopes are highly desirable. We evaluated Support Vector Machine (SVM) classifiers trained utilizing five different kernel methods using fivefold cross‐validation on a homology‐reduced data set of 701 linear B‐cell epitopes, extracted from Bcipep database, and 701 non‐epitopes, randomly extracted from SwissProt sequences. Based on the results of our computational experiments, we propose BCPred, a novel method for predicting linear B‐cell epitopes using the subsequence kernel. We show that the predictive performance of BCPred (AUC = 0.758) outperforms 11 SVM‐based classifiers developed and evaluated in our experiments as well as our implementation of AAP (AUC = 0.7), a recently proposed method for predicting linear B‐cell epitopes using amino acid pair antigenicity. Furthermore, we compared BCPred with AAP and ABCPred, a method that uses recurrent neural networks, using two data sets of unique B‐cell epitopes that had been previously used to evaluate ABCPred. Analysis of the data sets used and the results of this comparison show that conclusions about the relative performance of different B‐cell epitope prediction methods drawn on the basis of experiments using data sets of unique B‐cell epitopes are likely to yield overly optimistic estimates of performance of evaluated methods. This argues for the use of carefully homology‐reduced data sets in comparing B‐cell epitope prediction methods to avoid misleading conclusions about how different methods compare to each other. Our homology‐reduced data set and implementations of BCPred as well as the APP method are publicly available through our web‐based server, BCPREDS, at: http://ailab.cs.iastate.edu/bcpreds/. Copyright © 2008 John Wiley & Sons, Ltd.

Immune epitope database analysis resource (IEDB-AR)

We present a new release of the immune epitope database analysis resource (IEDB-AR, http://tools.immuneepitope.org), a repository of web-based tools for the prediction and analysis of immune epitopes. New functionalities have been added to most of the previously implemented tools, and a total of eight new tools were added, including two B-cell epitope prediction tools, four T-cell epitope prediction tools and two analysis tools.

The epitope space of the human proteome

In the post-genome era, there is a great need for protein-specific affinity reagents to explore the human proteome. Antibodies are suitable as reagents, but generation of antibodies with low cross-reactivity to other human proteins requires careful selection of antigens. Here we show the results from a proteome-wide effort to map linear epitopes based on uniqueness relative to the entire human proteome. The analysis was based on a sliding window sequence similarity search using short windows (8, 10, and 12 amino acid residues). A comparison of exact string matching (Hamming distance) and a heuristic method (BLAST) was performed, showing that the heuristic method combined with a grid strategy allows for whole proteome analysis with high accuracy and feasible run times. The analysis shows that it is possible to find unique antigens for a majority of the human proteins, with relatively strict rules involving low sequence identity of the possible linear epitopes. The implications for human antibody-based proteomics efforts are discussed.

Computer-aided biotechnology: from immuno-informatics to reverse vaccinology

Genome sequences from many organisms, including humans, have been completed, and high-throughput analyses have produced burgeoning volumes of 'omics' data. Bioinformatics is crucial for the management and analysis of such data and is increasingly used to accelerate progress in a wide variety of large-scale and object-specific functional analyses. Refined algorithms enable biotechnologists to follow 'computer-aided strategies' based on experiments driven by high-confidence predictions. In order to address compound problems, current efforts in immuno-informatics and reverse vaccinology are aimed at developing and tuning integrative approaches and user-friendly, automated bioinformatics environments. This will herald a move to 'computer-aided biotechnology': smart projects in which time-consuming and expensive large-scale experimental approaches are progressively replaced by prediction-driven investigations.

PEPOP: computational design of immunogenic peptides

Background

Most methods available to predict protein epitopes are sequence based. There is a need for methods using 3D information for prediction of discontinuous epitopes and derived immunogenic peptides.

Results

PEPOP uses the 3D coordinates of a protein both to predict clusters of surface accessible segments that might correspond to epitopes and to design peptides to be used to raise antibodies that target the cognate antigen at specific sites. To verify the ability of PEPOP to identify epitopes, 13 crystallographically defined epitopes were compared with PEPOP clusters: specificity ranged from 0.75 to 1.00, sensitivity from 0.33 to 1.00, and the positive predictive value from 0.19 to 0.89. Comparison of these results with those obtained with two other prediction algorithms showed comparable specificity and slightly better sensitivity and PPV. To prove the capacity of PEPOP to predict immunogenic peptides that induce protein cross-reactive antibodies, several peptides were designed from the 3D structure of model antigens (IA-2, TPO, and IL8) and chemically synthesized. The reactivity of the resulting anti-peptides antibodies with the cognate antigens was measured. In 80% of the cases (four out of five peptides), the flanking protein sequence process (sequence-based) of PEPOP successfully proposed peptides that elicited antibodies cross-reacting with the parent proteins. Polyclonal antibodies raised against peptides designed from amino acids which are spatially close in the protein, but separated in the sequence, could also be obtained, although they were much less reactive. The capacity of PEPOP to design immunogenic peptides that induce antibodies suitable for a sandwich capture assay was also demonstrated.

Conclusion

PEPOP has the potential to guide experimentalists that want to localize an epitope or design immunogenic peptides for raising antibodies which target proteins at specific sites. More successful predictions of immunogenic peptides were obtained when a peptide was continuous as compared with peptides corresponding to discontinuous epitopes. PEPOP is available for use at .

Analysis and prediction of protective continuous B-cell epitopes on pathogen proteins

Background

The application of peptide based diagnostics and therapeutics mimicking part of protein antigen is experiencing renewed interest. So far selection and design rationale for such peptides is usually driven by T-cell epitope prediction, available experimental and modelled 3D structure, B-cell epitope predictions such as hydrophilicity plots or experience. If no structure is available the rational selection of peptides for the production of functionally altering or neutralizing antibodies is practically impossible. Specifically if many alternative antigens are available the reduction of required synthesized peptides until one successful candidate is found is of central technical interest. We have investigated the integration of B-cell epitope prediction with the variability of antigen and the conservation of patterns for post-translational modification (PTM) prediction to improve over state of the art in the field. In particular the application of machine-learning methods shows promising results.

Results

We find that protein regions leading to the production of functionally altering antibodies are often characterized by a distinct increase in the cumulative sum of three presented parameters. Furthermore the concept to maximize antigenicity, minimize variability and minimize the likelihood of post-translational modification for the identification of relevant sites leads to biologically interesting observations. Primarily, for about 50% of antigen the approach works well with individual area under the ROC curve (AROC) values of at least 0.65. On the other hand a significant portion reveals equivalently low AROC values of < = 0.35 indicating an overall non-Gaussian distribution. While about a third of 57 antigens are seemingly intangible by our approach our results suggest the existence of at least two distinct classes of bioinformatically detectable epitopes which should be predicted separately. As a side effect of our study we present a hand curated dataset for the validation of protectivity classification. Based on this dataset machine-learning methods further improve predictive power to a class separation in an equilibrated dataset of up to 83%.

Conclusion

We present a computational method to automatically select and rank peptides for the stimulation of potentially protective or otherwise functionally altering antibodies. It can be shown that integration of variability, post-translational modification pattern conservation and B-cell antigenicity improve rational selection over random guessing. Probably more important, we find that for about 50% of antigen the approach works substantially better than for the overall dataset of 57 proteins. Essentially as a side effect our method optimizes for presumably best applicable peptides as they tend to be likely unmodified and as invariable as possible which is answering needs in diagnosis and treatment of pathogen infection. In addition we show the potential for further improvement by the application of machine-learning methods, in particular Random Forests.

Predicting flexible length linear B-cell epitopes

Identifying B-cell epitopes play an important role in vaccine design, immunodiagnostic tests, and antibody production. Therefore, computational tools for reliably predicting B-cell epitopes are highly desirable. We explore two machine learning approaches for predicting flexible length linear B-cell epitopes. The first approach utilizes four sequence kernels for determining a similarity score between any arbitrary pair of variable length sequences. The second approach utilizes four different methods of mapping a variable length sequence into a fixed length feature vector. Based on our empirical comparisons, we propose FBCPred, a novel method for predicting flexible length linear B-cell epitopes using the subsequence kernel. Our results demonstrate that FBCPred significantly outperforms all other classifiers evaluated in this study. An implementation of FBCPred and the datasets used in this study are publicly available through our linear B-cell epitope prediction server, BCPREDS, at: http://ailab.cs.iastate.edu/bcpreds/.

Modeling the adaptive immune system: predictions and simulations

MOTIVATION

Immunological bioinformatics methods are applicable to a broad range of scientific areas. The specifics of how and where they might be implemented have recently been reviewed in the literature. However, the background and concerns for selecting between the different available methods have so far not been adequately covered.

SUMMARY

Before using predictions systems, it is necessary to not only understand how the methods are constructed but also their strength and limitations. The prediction systems in humoral epitope discovery are still in their infancy, but have reached a reasonable level of predictive strength. In cellular immunology, MHC class I binding predictions are now very strong and cover most of the known HLA specificities. These systems work well for epitope discovery, and predictions of the MHC class I pathway have been further improved by integration with state-of-the-art prediction tools for proteasomal cleavage and TAP binding. By comparison, class II MHC binding predictions have not developed to a comparable accuracy level, but new tools have emerged that deliver significantly improved predictions not only in terms of accuracy, but also in MHC specificity coverage. Simulation systems and mathematical modeling are also now beginning to reach a level where these methods will be able to answer more complex immunological questions.

Immunoinformatics: an overview of computational tools and techniques for understanding immune function

In recent years, there has been a rapid expansion in the application of information technology to biological data. Although the use of information science techniques is less common for the discipline of immunology, this field has seen great strides in recent years. This review addresses why in silico modeling is needed in immunology research, highlights some of the major areas of research and suggests what may be important for the future of immunoinformatics.

Antibody-protein interactions: benchmark datasets and prediction tools evaluation

Background

The ability to predict antibody binding sites (aka antigenic determinants or B-cell epitopes) for a given protein is a precursor to new vaccine design and diagnostics. Among the various methods of B-cell epitope identification X-ray crystallography is one of the most reliable methods. Using these experimental data computational methods exist for B-cell epitope prediction. As the number of structures of antibody-protein complexes grows, further interest in prediction methods using 3D structure is anticipated. This work aims to establish a benchmark for 3D structure-based epitope prediction methods.

Results

Two B-cell epitope benchmark datasets inferred from the 3D structures of antibody-protein complexes were defined. The first is a dataset of 62 representative 3D structures of protein antigens with inferred structural epitopes. The second is a dataset of 82 structures of antibody-protein complexes containing different structural epitopes. Using these datasets, eight web-servers developed for antibody and protein binding sites prediction have been evaluated. In no method did performance exceed a 40% precision and 46% recall. The values of the area under the receiver operating characteristic curve for the evaluated methods were about 0.6 for ConSurf, DiscoTope, and PPI-PRED methods and above 0.65 but not exceeding 0.70 for protein-protein docking methods when the best of the top ten models for the bound docking were considered; the remaining methods performed close to random. The benchmark datasets are included as a supplement to this paper.

Conclusion

It may be possible to improve epitope prediction methods through training on datasets which include only immune epitopes and through utilizing more features characterizing epitopes, for example, the evolutionary conservation score. Notwithstanding, overall poor performance may reflect the generality of antigenicity and hence the inability to decipher B-cell epitopes as an intrinsic feature of the protein. It is an open question as to whether ultimately discriminatory features can be found.

Identification of discontinuous antigenic determinants on proteins based on shape complementarities

Diverse procedures for identifying antigenic determinants on proteins have been developed, including experimental as well as computational approaches. However, most of these techniques focus on continuous epitopes, whereas fast and reliable identification and verification of discontinuous epitopes remains barely amenable. In this paper, we describe a computational workflow for the detection of discontinuous epitopes on proteins. The workflow uses a given protein 3D structure as input, and combines a per residue solvent accessibility constraint with epitope to paratope shape complementarity measures and binding energies for assigning antigenic determinants in the conformational context. We have developed the procedure on a given set of 26 antigen-antibody complexes with a known structure, and have further expanded the available paratope shapes by generating a virtual paratope library in order to improve the screening for candidate residues constituting discontinuous epitopes. Applying the workflow on the 26 given antigens with known discontinuous epitopes resulted in the correct identification of the spatial proximity of 12 antigen-antibody interaction sites. Combining solvent accessibility, shape complementarity and binding energies towards the identification of discontinuous epitopes clearly outperforms approaches solely considering accessibility and residue distance constraints.

Towards a consensus on datasets and evaluation metrics for developing B-cell epitope prediction tools

A B-cell epitope is the three-dimensional structure within an antigen that can be bound to the variable region of an antibody. The prediction of B-cell epitopes is highly desirable for various immunological applications, but has presented a set of unique challenges to the bioinformatics and immunology communities. Improving the accuracy of B-cell epitope prediction methods depends on a community consensus on the data and metrics utilized to develop and evaluate such tools. A workshop, sponsored by the National Institute of Allergy and Infectious Disease (NIAID), was recently held in Washington, DC to discuss the current state of the B-cell epitope prediction field. Many of the currently available tools were surveyed and a set of recommendations was devised to facilitate improvements in the currently existing tools and to expedite future tool development. An underlying theme of the recommendations put forth by the panel is increased collaboration among research groups. By developing common datasets, standardized data formats, and the means with which to consolidate information, we hope to greatly enhance the development of B-cell epitope prediction tools.

Integrating epitope data into the emerging web of biomedical knowledge resources

In this age of information, keeping up with the literature can be overwhelming for any researcher. The generation of the IEBD – the first epitope-related database that makes complex and context-dependent information on immune epitopes readily accessible and searchable – may help

The recognition of immune epitopes is an important molecular mechanism of the vertebrate immune system to discriminate between self and non-self. Increasing amounts of data on immune epitopes are becoming available due to technological advances in epitope-mapping techniques and the availability of genomic information for pathogens. Organizing this data poses a challenge that is similar to the successful effort that was required to organize genomic data, which needed the establishment of centralized databases that complement the primary literature to make the data readily accessible and searchable by researchers. As described in this Innovation article, the Immune Epitope Database and Analysis Resource aims to achieve the same for the more complex and context-dependent information on immune epitopes, and to integrate this data with existing and emerging knowledge resources.

Immunoinformatics comes of age

With the burgeoning immunological data in the scientific literature, scientists must increasingly rely on Internet resources to inform and enhance their work. Here we provide a brief overview of the adaptive immune response and summaries of immunoinformatics resources, emphasizing those with Web interfaces. These resources include searchable databases of epitopes and immune-related molecules, and analysis tools for T cell and B cell epitope prediction, vaccine design, and protein structure comparisons. There is an agreeable synergy between the growing collections in immune-related databases and the growing sophistication of analysis software; the databases provide the foundation for developing predictive computational tools, which in turn enable more rapid identification of immune responses to populate the databases. Collectively, these resources contribute to improved understanding of immune responses and escape, and evolution of pathogens under immune pressure. The public health implications are vast, including designing vaccines, understanding autoimmune diseases, and defining the correlates of immune protection.

POPI: predicting immunogenicity of MHC class I binding peptides by mining informative physicochemical properties

MOTIVATION

Both modeling of antigen-processing pathway including major histocompatibility complex (MHC) binding and immunogenicity prediction of those MHC-binding peptides are essential to develop a computer-aided system of peptide-based vaccine design that is one goal of immunoinformatics. Numerous studies have dealt with modeling the immunogenic pathway but not the intractable problem of immunogenicity prediction due to complex effects of many intrinsic and extrinsic factors. Moderate affinity of the MHC-peptide complex is essential to induce immune responses, but the relationship between the affinity and peptide immunogenicity is too weak to use for predicting immunogenicity. This study focuses on mining informative physicochemical properties from known experimental immunogenicity data to understand immune responses and predict immunogenicity of MHC-binding peptides accurately.

RESULTS

This study proposes a computational method to mine a feature set of informative physicochemical properties from MHC class I binding peptides to design a support vector machine (SVM) based system (named POPI) for the prediction of peptide immunogenicity. High performance of POPI arises mainly from an inheritable bi-objective genetic algorithm, which aims to automatically determine the best number m out of 531 physicochemical properties, identify these m properties and tune SVM parameters simultaneously. The dataset consisting of 428 human MHC class I binding peptides belonging to four classes of immunogenicity was established from MHCPEP, a database of MHC-binding peptides (Brusic et al., 1998). POPI, utilizing the m = 23 selected properties, performs well with the accuracy of 64.72% using leave-one-out cross-validation, compared with two sequence alignment-based prediction methods ALIGN (54.91%) and PSI-BLAST (53.23%). POPI is the first computational system for prediction of peptide immunogenicity based on physicochemical properties.

AVAILABILITY

A web server for prediction of peptide immunogenicity (POPI) and the used dataset of MHC class I binding peptides (PEPMHCI) are available at http://iclab.life.nctu.edu.tw/POPI

Cryptic B-cell epitope identification through informational analysis of protein sequenses

A comparison of the location of B-cell epitopes and information structure (IS) of protein sequences was attempted. Analysis of 62 known B-cell epitopes located in five different proteins showed that they concentrated in IS sites with increased degree of information coordination. Based on the analysis of IS six peptides from two proteins were selected and produced in a recombinant form as yeast virus-like particles (VLPs). Immunization of mice with recombinant VLP-peptides has induced the production of IgG capable of recognizing full-length antigens. This result suggests that the analysis of IS of proteins can be useful in the selection of peptides possessing cryptic B-cell epitope activity.

Prediction of linear B-cell epitopes using amino acid pair antigenicity scale

Identification of antigenic sites on proteins is of vital importance for developing synthetic peptide vaccines, immunodiagnostic tests and antibody production. Currently, most of the prediction algorithms rely on amino acid propensity scales using a sliding window approach. These methods are oversimplified and yield poor predicted results in practice. In this paper, a novel scale, called the amino acid pair (AAP) antigenicity scale, is proposed that is based on the finding that B-cell epitopes favor particular AAPs. It is demonstrated that, using SVM (support vector machine) classifier, the AAP antigenicity scale approach has much better performance than the existing scales based on the single amino acid propensity. The AAP antigenicity scale can reflect some special sequence-coupled feature in the B-cell epitopes, which is the essence why the new approach is superior to the existing ones. It is anticipated that with the continuous increase of the known epitope data, the power of the AAP antigenicity scale approach will be further enhanced.

Prediction methods for B-cell epitopes

In this chapter, two prediction servers of linear B-cell epiotpes have been described; (i) BcePred, based on physico-chemical properties that include hydrophilicity, flexibility/mobility, accessibility, polarity, exposed surface, turns, and antigenicity and ii) ABCpred, based on recurrent neural network. Both of the servers assist in locating linear epitope regions in a protein.

How reliable is the structural prediction of IgE-binding epitopes of allergens? The case study of plant lipid transfer proteins

The linear IgE-binding epitopes of non-specific lipid transfer proteins (nsLTP) from plants were predicted using a combination of predictive tools including (1) the hydropathic profiles based on different scales of hydrophilicity, flexibility and exposure to the solvent, (2) the hydrophobic cluster analysis plots, (3) the occurrence of charged residues in the predicted amino acid sequence stretches and, (4) the exposition of the predicted linear IgE-binding epitopes checked on the three-dimensional models built for the nsLTP. A reliable prediction was obtained for nsLTP as compared with the previously characterized IgE-binding epitopes of various proteins. A consensual IgE-binding epitope occurring in other plant nsLTP and responsible for some IgE-binding cross-reactivity among fruit nsLTP has been identified and characterized. Despite some discrepancies, a fairly good prediction resulted in applying our combination of predictive methods to longer nsLTP or plant profilins.

Prediction of residues in discontinuous B-cell epitopes using protein 3D structures

Discovery of discontinuous B-cell epitopes is a major challenge in vaccine design. Previous epitope prediction methods have mostly been based on protein sequences and are not very effective. Here, we present DiscoTope, a novel method for discontinuous epitope prediction that uses protein three-dimensional structural data. The method is based on amino acid statistics, spatial information, and surface accessibility in a compiled data set of discontinuous epitopes determined by X-ray crystallography of antibody/antigen protein complexes. DiscoTope is the first method to focus explicitly on discontinuous epitopes. We show that the new structure-based method has a better performance for predicting residues of discontinuous epitopes than methods based solely on sequence information, and that it can successfully predict epitope residues that have been identified by different techniques. DiscoTope detects 15.5% of residues located in discontinuous epitopes with a specificity of 95%. At this level of specificity, the conventional Parker hydrophilicity scale for predicting linear B-cell epitopes identifies only 11.0% of residues located in discontinuous epitopes. Predictions by the DiscoTope method can guide experimental epitope mapping in both rational vaccine design and development of diagnostic tools, and may lead to more efficient epitope identification.

Selection of cryptic B-cell epitopes using informational analysis of protein sequences

Sub-unit vaccines are synthetic or recombinant peptides representing T- or B-cell epitopes of major protein antigens from a particular pathogen. Epitope selection requires the synthesis of peptides that overlap the protein sequences and screening for the most effective ones. In this study a new method of immunogenic peptide selection based on the analysis of information structure of protein sequences is suggested. The analysis of known B-cell epitope location in the information structure of Aspergillus fumigatus proteins Asp f 2 and Asp f 3 has shown that epitopes are scattered along the sequences of proteins for the exception of sites with Increased Degree Information Coordination (IDIC). Based on these results peptides from different allergens such as Asp f 2, Der p 1, and Fel d 1 were selected and produced in a recombinant form in the context of yeast virus-like particles (VLPs). Immunization of mice with VLPs containing peptides form allergens has induced the production of IgG able to recognize full-length antigens. This result suggests that the analysis of information structure of proteins can be used for the selection of peptides possessing cryptic B-cell epitope activity.

AlgPred: prediction of allergenic proteins and mapping of IgE epitopes

In this study a systematic attempt has been made to integrate various approaches in order to predict allergenic proteins with high accuracy. The dataset used for testing and training consists of 578 allergens and 700 non-allergens obtained from A. K. Bjorklund, D. Soeria-Atmadja, A. Zorzet, U. Hammerling and M. G. Gustafsson (2005) Bioinformatics, 21, 39-50. First, we developed methods based on support vector machine using amino acid and dipeptide composition and achieved an accuracy of 85.02 and 84.00%, respectively. Second, a motif-based method has been developed using MEME/MAST software that achieved sensitivity of 93.94 with 33.34% specificity. Third, a database of known IgE epitopes was searched and this predicted allergenic proteins with 17.47% sensitivity at specificity of 98.14%. Fourth, we predicted allergenic proteins by performing BLAST search against allergen representative peptides. Finally hybrid approaches have been developed, which combine two or more than two approaches. The performance of all these algorithms has been evaluated on an independent dataset of 323 allergens and on 101 725 non-allergens obtained from Swiss-Prot. A web server AlgPred has been developed for the predicting allergenic proteins and for mapping IgE epitopes on allergenic proteins (http://www.imtech.res.in/raghava/algpred/). AlgPred is available at www.imtech.res.in/raghava/algpred/.

Prediction of continuous B-cell epitopes in an antigen using recurrent neural network

B-cell epitopes play a vital role in the development of peptide vaccines, in diagnosis of diseases, and also for allergy research. Experimental methods used for characterizing epitopes are time consuming and demand large resources. The availability of epitope prediction method(s) can rapidly aid experimenters in simplifying this problem. The standard feed-forward (FNN) and recurrent neural network (RNN) have been used in this study for predicting B-cell epitopes in an antigenic sequence. The networks have been trained and tested on a clean data set, which consists of 700 non-redundant B-cell epitopes obtained from Bcipep database and equal number of non-epitopes obtained randomly from Swiss-Prot database. The networks have been trained and tested at different input window length and hidden units. Maximum accuracy has been obtained using recurrent neural network (Jordan network) with a single hidden layer of 35 hidden units for window length of 16. The final network yields an overall prediction accuracy of 65.93% when tested by fivefold cross-validation. The corresponding sensitivity, specificity, and positive prediction values are 67.14, 64.71, and 65.61%, respectively. It has been observed that RNN (JE) was more successful than FNN in the prediction of B-cell epitopes. The length of the peptide is also important in the prediction of B-cell epitopes from antigenic sequences. The webserver ABCpred is freely available at www.imtech.res.in/raghava/abcpred/.

Machine learning approaches for prediction of linear B-cell epitopes on proteins

Identification and characterization of antigenic determinants on proteins has received considerable attention utilizing both, experimental as well as computational methods. For computational routines mostly structural as well as physicochemical parameters have been utilized for predicting the antigenic propensity of protein sites. However, the performance of computational routines has been low when compared to experimental alternatives. Here we describe the construction of machine learning based classifiers to enhance the prediction quality for identifying linear B-cell epitopes on proteins. Our approach combines several parameters previously associated with antigenicity, and includes novel parameters based on frequencies of amino acids and amino acid neighborhood propensities. We utilized machine learning algorithms for deriving antigenicity classification functions assigning antigenic propensities to each amino acid of a given protein sequence. We compared the prediction quality of the novel classifiers with respect to established routines for epitope scoring, and tested prediction accuracy on experimental data available for HIV proteins. The major finding is that machine learning classifiers clearly outperform the reference classification systems on the HIV epitope validation set.

Selection and combination of machine learning classifiers for prediction of linear B-cell epitopes on proteins

Recently, new machine learning classifiers for the prediction of linear B-cell epitopes were presented. Here we show the application of Receiver Operator Characteristics (ROC) convex hulls to select optimal classifiers as well as possibilities to improve the post test probability (PTP) to meet real world requirements such as high throughput epitope screening of whole proteomes. The major finding is that ROC convex hulls present an easy to use way to rank classifiers based on their prediction conservativity as well as to select candidates for ensemble classifiers when validating against the antigenicity profile of 10 HIV-1 proteins. We also show that linear models are at least equally efficient to model the available data when compared to multi-layer feed-forward neural networks.

Improved method for predicting linear B-cell epitopes

BACKGROUND

B-cell epitopes are the sites of molecules that are recognized by antibodies of the immune system. Knowledge of B-cell epitopes may be used in the design of vaccines and diagnostics tests. It is therefore of interest to develop improved methods for predicting B-cell epitopes. In this paper, we describe an improved method for predicting linear B-cell epitopes.

RESULTS

In order to do this, three data sets of linear B-cell epitope annotated proteins were constructed. A data set was collected from the literature, another data set was extracted from the AntiJen database and a data sets of epitopes in the proteins of HIV was collected from the Los Alamos HIV database. An unbiased validation of the methods was made by testing on data sets on which they were neither trained nor optimized on. We have measured the performance in a non-parametric way by constructing ROC-curves.

CONCLUSION

The best single method for predicting linear B-cell epitopes is the hidden Markov model. Combining the hidden Markov model with one of the best propensity scale methods, we obtained the BepiPred method. When tested on the validation data set this method performs significantly better than any of the other methods tested. The server and data sets are publicly available at http://www.cbs.dtu.dk/services/BepiPred.