Allergenicity prediction by protein sequence

PreAlgPro: Prediction of allergenic proteins with pre-trained protein language model and efficient neutral network

Allergy is a prevalent phenomenon, involving allergens such as nuts and milk. Avoiding exposure to allergens is the most effective preventive measure against allergic reactions. However, current homology-based methods for identifying allergenic proteins encounter challenges when dealing with non-homologous data. Traditional machine learning approaches rely on manually extracted features, which lack important protein functional characteristics, including evolutionary information. Consequently, there is still considerable room for improvement in existing methods. In this study, we present PreAlgPro, a method for identifying allergenic proteins based on pre-trained protein language models and deep learning techniques. Specifically, we employed the ProtT5 model to extract protein embedding features, replacing the manual feature extraction step. Furthermore, we devised an Attention-CNN neural network architecture to identify potential features that contribute to the classification of allergenic proteins. The performance of our model was evaluated on four independent test sets, and the experimental results demonstrate that PreAlgPro surpasses existing state-of-the-art methods. Additionally, we collected allergenic protein samples to validate the robustness of the model and conducted an analysis of model interpretability.

Bioinformatic and literature assessment of toxicity and allergenicity of a CRISPR-Cas9 engineered gene drive to control Anopheles gambiae the mosquito vector of human malaria

BACKGROUND

Population suppression gene drive is currently being evaluated, including via environmental risk assessment (ERA), for malaria vector control. One such gene drive involves the dsxFCRISPRh transgene encoding (i) hCas9 endonuclease, (ii) T1 guide RNA (gRNA) targeting the doublesex locus, and (iii) DsRed fluorescent marker protein, in genetically-modified mosquitoes (GMMs). Problem formulation, the first stage of ERA, for environmental releases of dsxFCRISPRh previously identified nine potential harms to the environment or health that could occur, should expressed products of the transgene cause allergenicity or toxicity.

METHODS

Amino acid sequences of hCas9 and DsRed were interrogated against those of toxins or allergens from NCBI, UniProt, COMPARE and AllergenOnline bioinformatic databases and the gRNA was compared with microRNAs from the miRBase database for potential impacts on gene expression associated with toxicity or allergenicity. PubMed was also searched for any evidence of toxicity or allergenicity of Cas9 or DsRed, or of the donor organisms from which these products were originally derived.

RESULTS

While Cas9 nuclease activity can be toxic to some cell types in vitro and hCas9 was found to share homology with the prokaryotic toxin VapC, there was no evidence from previous studies of a risk of toxicity to humans and other animals from hCas9. Although hCas9 did contain an 8-mer epitope found in the latex allergen Hev b 9, the full amino acid sequence of hCas9 was not homologous to any known allergens. Combined with a lack of evidence in the literature of Cas9 allergenicity, this indicated negligible risk to humans of allergenicity from hCas9. No matches were found between the gRNA and microRNAs from either Anopheles or humans. Moreover, potential exposure to dsxFCRISPRh transgenic proteins from environmental releases was assessed as negligible.

CONCLUSIONS

Bioinformatic and literature assessments found no convincing evidence to suggest that transgenic products expressed from dsxFCRISPRh were allergens or toxins, indicating that environmental releases of this population suppression gene drive for malaria vector control should not result in any increased allergenicity or toxicity in humans or animals. These results should also inform evaluations of other GMMs being developed for vector control and in vivo clinical applications of CRISPR-Cas9.

DeepAlgPro: an interpretable deep neural network model for predicting allergenic proteins

Allergies have become an emerging public health problem worldwide. The most effective way to prevent allergies is to find the causative allergen at the source and avoid re-exposure. However, most of the current computational methods used to identify allergens were based on homology or conventional machine learning methods, which were inefficient and still had room to be improved for the detection of allergens with low homology. In addition, few methods based on deep learning were reported, although deep learning has been successfully applied to several tasks in protein sequence analysis. In the present work, a deep neural network-based model, called DeepAlgPro, was proposed to identify allergens. We showed its great accuracy and applicability to large-scale forecasts by comparing it to other available tools. Additionally, we used ablation experiments to demonstrate the critical importance of the convolutional module in our model. Moreover, further analyses showed that epitope features contributed to model decision-making, thus improving the model's interpretability. Finally, we found that DeepAlgPro was capable of detecting potential new allergens. Overall, DeepAlgPro can serve as powerful software for identifying allergens.

Novel Machine Learning Method AllerStat Identifies Statistically Significant Allergen-Specific Patterns in Protein Sequences

Cutting-edge technologies such as genome editing and synthetic biology allow us to produce novel foods and functional proteins. However, their toxicity and allergenicity must be accurately evaluated. It is known that specific amino-acid sequences in proteins make some proteins allergic, but many of these sequences remain uncharacterized. In this study, we introduce a data-driven approach and a machine-learning (ML) method to find undiscovered allergen specific patterns (ASPs) among amino acid sequences. The proposed method enables an exhaustive search for amino-acid subsequences whose frequencies are statistically significantly higher in allergenic proteins. As a proof-of-concept, we created a database containing 21,154 proteins of which the presence or absence of allergic reactions are already known, and applied the proposed method to the database. The detected ASPs in this proof-of-concept study were consistent with known biological findings, and the allergenicity prediction performance using the detected ASPs was higher than extant approaches, indicating this method may be useful in evaluating the utility of synthetic foods and proteins.

Herbivory-inducible lipid-transfer proteins (LTPs) of Cicer arietinum as potential human allergens

Lipid-transfer proteins (LTPs) are lipid-binding small proteins, ubiquitously distributed amongst plant kingdom. Apart from their involvement in plant defense, it has also been discovered that they induce allergic reactions in humans. A plethora of LTPs have been identified in vegetables, fruits, pollens, nuts, and latex, among which Pru p 3, a LTP allergen from peach fruit, is extensively studied and exhibits cross-reactivity with potential allergens from different species. In Cicer arietinum, a family of LTPs (CaLTPs) has been identified and their importance in plant defense during Helicoverpa armigera-infestation has been recognized. However, the determination of the allergenicity potential of CaLTPs has not been attempted. In this study, we aim to decipher the allergenicity potential of defense-related CaLTPs. The allergenicity potential prediction, and identification of B-cell epitope binding regions showed that the CaLTPs had conserved domains and B-cell epitopes in the same regions as Prup3 (a marker allergen for LTPs). Using molecular docking and simulations, we observed that the CaLTPs successfully interacted with the Immunoglobin E(IgE)with docking energies ranging from -315.5 to -268.4 and the structures were stabilized within 10 ns of simulation. Through this study, we intend to embellish our present knowledge and understanding of the sensitization and allergenicity potential of CaLTPs.Communicated by Ramaswamy H. Sarma.

Potential of a novel flagellin epitope as a broad-spectrum vaccine candidate against enteric fever

Relentless emergence of antibiotic resistant Salmonella strains, coupled with the drawbacks associated with currently available vaccines against enteric fever, warrants an urgent need to look for new vaccine candidates. Out of the multiple virulence factors harbored by Salmonella, flagella are regarded as one of the most important targets of innate as well as adaptive immune response. Individual Salmonella serotypes alternate between expression of two different antigenic forms encoded by fliC and fljB genes, respectively thereby employing this as a strategy to escape the host immune response. In the present study, using various immunoinformatic approaches, a flagellin epitope, present in both antigenic forms of typhoidal Salmonellae has been targeted. Following B-cell epitope and B-cell derived T-cell epitope prediction and interaction studies with major histocompatibility complexes using molecular docking, a peptide epitope was selected. Further, it was screened for its presence in majority of typhoidal serovars along with other useful attributes, in silico. Thereafter, safety studies were performed with the synthesized peptide. Subsequently, immunization studies were carried out using S. Typhi as well as S. Paratyphi A induced murine peritonitis model. Active immunization with peptide-BSA conjugate resulted in 75% and 80% mice survival following lethal challenge with S. Typhi and S. Paratyphi A respectively, along with a significant IgG antibody titer, thereby highlighting its immunogenic potential. Reduced bacterial burden in vital organs along with improved histoarchitecture and cytokine levels further substantiated the protective efficacy of the proposed candidate. Passive immunization studies with the candidate verified the protective efficacy of the generated antibodies against lethal challenge of bacteria in mice. Given the endemic nature of enteric fever and the antigenic variability observed in Salmonella serotypes, present study highlights the importance of using a vaccine candidate, which, along with generating a strong immune response, also exhibits a broad coverage against both, S. Typhi as well as S. Paratyphi A strains.

Development of epitope-based chimeric protein as a vaccine against Lujo virus by utilizing immunoinformatic tools

Aim: Lujo is a modern zoonotic virus that is potentially fatal and spreads by bodily fluids. In this research, immunoinformatic tools are used to build a vaccine. Methodology: The epitopes of cytotoxic T-lymphocytes, helper T-lymphocytes and linear B-lymphocytes were predicted from the most antigenic protein. The designed vaccine's physiochemical properties and 3D structure have been forecasted. Low free energy and strong binding affinity estimated in molecular docking against toll-like receptor 4 (TLR4) and dynamic simulation. Furthermore, in silico cloning in the Escherichia coli K12 host system was performed for high level of expression. Conclusion: Finally, immune simulation was used to determine immune responses to the vaccine that was formulated confirming the developed vaccine as a good candidate against Lujo virus.

A Combined Proteomic and Metabolomic Strategy for Allergens Characterization in Natural and Fermented Brassica napus Bee Pollen

Bee pollen is consumed for its nutritional and pharmacological benefits, but it also contains hazardous allergens which have not been identified. Here, we identified two potential allergens, glutaredoxin and oleosin-B2, in Brassica napus bee pollen using mass spectrometry-based proteomics analyses, and used bioinformatics to predict their antigenic epitopes. Comparison of fermented (by Saccharomyces cerevisiae) and unfermented bee pollen samples indicated that glutaredoxin and oleosin-B2 contents were significantly decreased following fermentation, while the contents of their major constituent oligopeptides and amino acids were significantly increased based on metabolomics analyses. Immunoblot analysis indicated that the IgE-binding affinity with extracted bee pollen proteins was also significantly decreased after fermentation, suggesting a reduction in the allergenicity of fermented bee pollen. Furthermore, fermentation apparently promoted the biosynthesis of L-valine, L-isoleucine, L-tryptophan, and L-phenylalanine, as well as their precursors or intermediates. Thus, fermentation could potentially alleviate allergenicity, while also positively affecting nutritional properties of B. napus bee pollen. Our findings might provide a scientific foundation for improving the safety of bee pollen products to facilitate its wider application.

Allergic Diseases: A Comprehensive Review on Risk Factors, Immunological Mechanisms, Link with COVID-19, Potential Treatments, and Role of Allergen Bioinformatics

The prevalence of allergic diseases is regarded as one of the key challenges in health worldwide. Although the precise mechanisms underlying this rapid increase in prevalence are unknown, emerging evidence suggests that genetic and environmental factors play a significant role. The immune system, microbiota, viruses, and bacteria have all been linked to the onset of allergy disorders in recent years. Avoiding allergen exposure is the best treatment option; however, steroids, antihistamines, and other symptom-relieving drugs are also used. Allergen bioinformatics encompasses both computational tools/methods and allergen-related data resources for managing, archiving, and analyzing allergological data. This study highlights allergy-promoting mechanisms, algorithms, and concepts in allergen bioinformatics, as well as major areas for future research in the field of allergology.

Design of Monovalent and Chimeric Tetravalent Dengue Vaccine Using an Immunoinformatics Approach

An immunoinformatics technique was used to predict a monovalent amide immunogen candidate capable of producing therapeutic antibodies as well as a potent immunogen candidate capable of acting as a universal vaccination against all dengue fever virus serotypes. The capsid protein is an attractive goal for anti-DENV due to its position in the dengue existence cycle. The widely accessible immunological data, advances in antigenic peptide prediction using reverse vaccinology, and the introduction of molecular docking in immunoinformatics have directed vaccine manufacturing. The C-proteins of DENV-1-4 serotypes were known as antigens to assist with logical design. Binding epitopes for TC cells, TH cells, and B cells is predicted from structural dengue virus capsid proteins. Each T cell epitope of C-protein integrated with a B cell as a templet was used as a vaccine and produce antibodies in contrast to serotype of the dengue virus. A chimeric tetravalent vaccine was created by combining four vaccines, each representing four dengue serotypes, to serve as a standard vaccine candidate for all four Sero groups. The LKRARNRVS, RGFRKEIGR, KNGAIKVLR, and KAINVLRGF from dengue 1, dengue 2, dengue 3, and dengue 4 epitopes may be essential immunotherapeutic representatives for controlling outbreaks.

New applications of advanced instrumental techniques for the characterization of food allergenic proteins

Current approaches based on electrophoretic, chromatographic or immunochemical principles have allowed characterizing multiple allergens, mapping their epitopes, studying their mechanisms of action, developing detection and diagnostic methods and therapeutic strategies for the food and pharmaceutical industry. However, some of the common structural features related to the allergenic potential of food proteins remain unknown, or the pathological mechanism of food allergy is not yet fully understood. In addition, it is also necessary to evaluate new allergens from novel protein sources that may pose a new risk for consumers. Technological development has allowed the expansion of advanced technologies for which their whole potential has not been entirely exploited and could provide novel contributions to still unexplored molecular traits underlying both the structure of food allergens and the mechanisms through which they sensitize or elicit adverse responses in human subjects, as well as improving analytical techniques for their detection. This review presents cutting-edge instrumental techniques recently applied when studying structural and functional aspects of proteins, mechanism of action and interaction between biomolecules. We also exemplify their role in the food allergy research and discuss their new possible applications in several areas of the food allergy field.

Sequence-based Identification of Allergen Proteins Developed by Integration of PseAAC and Statistical Moments via 5-Step Rule

Background:

Allergens are antigens that can stimulate an atopic type I human hypersensitivity reaction by an immunoglobulin E (IgE) reaction. Some proteins are naturally allergenic than others. The challenge for toxicologists is to identify properties that allow proteins to cause allergic sensitization and allergic diseases. The identification of allergen proteins is a very critical and pivotal task. The experimental identification of protein functions is a hectic, laborious and costly task; therefore, computer scientists have proposed various methods in the field of computational biology and bioinformatics using various data science approaches. Objectives: Herein, we report a novel predictor for the identification of allergen proteins.

Methods:

For feature extraction, statistical moments and various position-based features have been incorporated into Chou’s pseudo amino acid composition (PseAAC), and are used for training of a neural network.

Results:

The predictor is validated through 10-fold cross-validation and Jackknife testing, which gave 99.43% and 99.87% accurate results.

Conclusions:

Thus, the proposed predictor can help in predicting the Allergen proteins in an efficient and accurate way and can provide baseline data for the discovery of new drugs and biomarkers.

An overview of in silico vaccine design against different pathogens and cancer

INTRODUCTION

Due to overcome the hardness of the vaccine design, computational vaccinology is emerging widely. Prediction of T cell and B cell epitopes, antigen processing analysis, antigenicity analysis, population coverage, conservancy analysis, allergenicity assessment, toxicity prediction, and protein-peptide docking are important steps in the process of designing and developing potent vaccines against various viruses and cancers. In order to perform all of the analyses, several bioinformatics tools and online web servers have been developed. Scientists must take the decision to apply more suitable and precise servers for each part based on their accuracy.

AREAS COVERED

In this review, a wide-range list of different bioinformatics tools and online web servers has been provided. Moreover, some studies were proposed to show the importance of various bioinformatics tools for predicting and developing efficient vaccines against different pathogens including viruses, bacteria, parasites, and fungi as well as cancer.

EXPERT OPINION

Immunoinformatics is the best way to find potential vaccine candidates against different pathogens. Thus, the selection of the most accurate tools is necessary to predict and develop potent preventive and therapeutic vaccines. To further evaluation of the computational and in silico vaccine design, in vitro/in vivo analyses are required to develop vaccine candidates.

AllerCatPro-prediction of protein allergenicity potential from the protein sequence

Motivation

Due to the risk of inducing an immediate Type I (IgE-mediated) allergic response, proteins intended for use in consumer products must be investigated for their allergenic potential before introduction into the marketplace. The FAO/WHO guidelines for computational assessment of allergenic potential of proteins based on short peptide hits and linear sequence window identity thresholds misclassify many proteins as allergens.

Results

We developed AllerCatPro which predicts the allergenic potential of proteins based on similarity of their 3D protein structure as well as their amino acid sequence compared with a data set of known protein allergens comprising of 4180 unique allergenic protein sequences derived from the union of the major databases Food Allergy Research and Resource Program, Comprehensive Protein Allergen Resource, WHO/International Union of Immunological Societies, UniProtKB and Allergome. We extended the hexamer hit rule by removing peptides with high probability of random occurrence measured by sequence entropy as well as requiring 3 or more hexamer hits consistent with natural linear epitope patterns in known allergens. This is complemented with a Gluten-like repeat pattern detection. We also switched from a linear sequence window similarity to a B-cell epitope-like 3D surface similarity window which became possible through extensive 3D structure modeling covering the majority (74%) of allergens. In case no structure similarity is found, the decision workflow reverts to the old linear sequence window rule. The overall accuracy of AllerCatPro is 84% compared with other current methods which range from 51 to 73%. Both the FAO/WHO rules and AllerCatPro achieve highest sensitivity but AllerCatPro provides a 37-fold increase in specificity.

Availability and implementation

https://allercatpro.bii.a-star.edu.sg/

Supplementary information

Supplementary data are available at Bioinformatics online.

Extensive immunoinformatics study for the prediction of novel peptide-based epitope vaccine with docking confirmation against envelope protein of Chikungunya virus: a computational biology approach

Chikungunya virus (CHIKV) instigating Chikungunya fever is a global infective menace resulting in high fever, weakened joint-muscle pain, and brain inflammation. Inaccessibility and unavailability of effective drugs have led us to an uncertain arena when it comes to providing proper medical treatment to the affected people. In this study, authentic encroachment has been made concerning the peptide-based epitope vaccine designing against CHIKV. A Proteome-wide search was performed to locate a conserved portion among the accessible viral outer membrane proteins which showcase a remarkable immune response using specific immunoinformatics and docking simulation tools. Primarily, the most probable immunogenic envelope glycoproteins E1 and E2 were identified from the UniProt database depending on their antigenicity scores. Subsequently, we selected two distinctive sequences "SEDVYANTQLVLQRP" and "IMLLYPDHPTLLSYR" in both E1 and E2 glycoproteins respectively. These two sequences identified as the most potent T and B cell epitope-based peptides as they interacted with 6 and 7 HLA-I and 5 HLA-II molecules with an extremely low IC50 score that was verified by molecular docking. Moreover, the sequences possess no allergenicity and are certainly located outside the transmembrane region. In addition, the sequences exhibited 88.46% and 100.00% Conservancy, covering high population coverage of 89.49% to 94.74% and 60.51% to 88.87% respectively in endemic countries. The identified peptide SEDVYANTQLVLQRP and IMLLYPDHPTLLSYR can be utilized next for the development of peptide-based epitope vaccine contrary to CHIKV, so further documentations and experimentations like Antigen testing, Antigen production, Clinical trials are needed to prove the validity of it. Communicated by Ramaswamy H. Sarma.

Identification and characterization of a sterically robust phenylalanine ammonia-lyase among 481 natural isoforms through association of in silico and in vitro studies

The enzyme phenylalanine ammonia lyase (PAL) is of special importance for the treatment of phenylketonuria patients. The aim of this study was to find a stable recombinant PAL with suitable kinetic properties among all natural PAL producing species using in silico and experimental approaches. To find such a stable PAL among 481 natural isoforms, 48,000 of 3-D models were predicted using the Modeller 9.10 program and evaluated by Ramachandran plot. Correlation analysis between Ramachandran plot and the energy of different thermodynamic components indicated that this plot could be an appropriate tool to predict protein stability. Hence, PAL6 from Lotus japonicus (LjPAL6) was selected as a stable isoform. Molecular dynamic (MD) simulation for 50 ns and docking has been conducted for LjPAL6-phenylalanine complex. The best PAL-phenylalanine frame was selected by re-docking with l-phenylalanine (L-Phe) and root-mean-square deviation (RMSD) value. MD simulation showed that the complex has a good stability, depicted by the low RMSD value, binding free energy and hydrogen bindings. Docking results showed that LjPAL6 has a high affinity toward l-Phe according to the low level of binding free energy. By overexpressing Ljpal6 in E. coli BL21, a total of 33.5 mg/l of protein was obtained, which has been increased to 83.7 mg/l via the optimization of LjPAL6 production using response surface methodology. The optimal pH and temperature were 8.5 and 50 °C, respectively. LjPAL6 showed a specific activity of 42 nkat/mg protein, with K_m, K_cat and K_cat/K_m values of 0.483 mM, 7 S^-1 and 14.5 S^-1 mM^-1 for l-phe, respectively. In conclusion, finding models with the most reasonable stereo-chemical quality and lowest numbers of steric clashes would result in easier folding. Hence, in silico analyses of bulk data from natural origin will lead one to find an optimal model for in vitro studies and drug design.

Assessment of the potential allergenicity of genetically-engineered food crops

An extensive safety assessment process exists for genetically-engineered (GE) crops. The assessment includes an evaluation of the introduced protein as well as the crop containing the protein with the goal of demonstrating the GE crop is "as-safe-as" non-GE crops in the food supply. One of the evaluations for GE crops is to assess the expressed protein for allergenic potential. Currently, no single factor is recognized as a predictor for protein allergenicity. Therefore, a weight-of-the-evidence approach, which accounts for a variety of factors and approaches for an overall assessment of allergenic potential, is conducted. This assessment includes an evaluation of the history of exposure and safety of the gene(s) source; protein structure (e.g. amino acid sequence identity to human allergens); stability of the protein to pepsin digestion in vitro; heat stability of the protein; glycosylation status; and when appropriate, specific IgE binding studies with sera from relevant clinically allergic subjects. Since GE crops were first commercialized over 20 years ago, there is no proof that the introduced novel protein(s) in any commercialized GE food crop has caused food allergy.

Q-X1-P-X2 motif search for potential celiac disease risk has poor selectivity

The European Food Safety Authority (EFSA) recently published guidelines for assessment of potential celiac disease risk for newly expressed proteins in genetically modified (GM) crops. This novel step-wise approach prescribes, in part, how to conduct sequence identity searches between a newly expressed protein and known celiac disease peptides including a Q/E-X1-P-X2 amino acid motif. To evaluate the specificity of the recommended sequence identity searches in the context of risk assessment, protein sequences from celiac disease causing crops, as well as from crops not associated with celiac disease, were compared with known HLA-DQ restricted epitopes and searched for the presence of motifs followed by peptide analysis. Searches for the presence of the Q/E-X1-P-X2-motif were found to generate a high proportion of false-positive hits irrelevant to celiac disease risk. Identification of a 9mer exact match between a newly expressed protein and the known celiac disease peptides (recommended by the guideline) along with a supplementary sequence comparisons (suggested by FARRP/AllergenOnline) is considered better suited to more specifically capture the potential risk of a newly expressed protein for celiac disease.

Distinguishing allergens from non-allergenic homologues using Physical-Chemical Property (PCP) motifs

Quantitative guidelines to distinguish allergenic proteins from related, but non-allergenic ones are urgently needed for regulatory agencies, biotech companies and physicians. In a previous study, we found that allergenic proteins populate a relatively small number of protein families, as characterized by the Pfam database. However, these families also contain non-allergenic proteins, meaning that allergenic determinants must lie within more discrete regions of the sequence. Thus, new methods are needed to discriminate allergenic proteins within those families. Physical-Chemical Properties (PCP)-motifs specific for allergens within a Pfam class were determined for 17 highly populated protein domains. A novel scoring method based on PCP-motifs that characterize known allergenic proteins within these families was developed, and validated for those domains. The motif scores distinguished sequences of allergens from a large selection of 80,000 randomly selected non-allergenic sequences. The motif scores for the birch pollen allergen (Bet v 1) family, which also contains related fruit and nut allergens, correlated better than global sequence similarities with clinically observed cross-reactivities among those allergens. Further, we demonstrated that the average scores of allergen specific motifs for allergenic profilins are significantly different from the scores of non-allergenic profilins. Several of the selective motifs coincide with experimentally determined IgE epitopes of allergenic profilins. The motifs also discriminated allergenic pectate lyases, including Jun a 1 from mountain cedar pollen, from similar proteins in the human microbiome, which can be assumed to be non-allergens. The latter lacked key motifs characteristic of the known allergens, some of which correlate with known IgE binding sites.

Improving Allergen Prediction in Main Crops Using a Weighted Integrative Method

As a public health problem, food allergy is frequently caused by food allergy proteins, which trigger a type-I hypersensitivity reaction in the immune system of atopic individuals. The food allergens in our daily lives are mainly from crops including rice, wheat, soybean and maize. However, allergens in these main crops are far from fully uncovered. Although some bioinformatics tools or methods predicting the potential allergenicity of proteins have been proposed, each method has their limitation. In this paper, we built a novel algorithm PREALW, which integrated PREAL, FAO/WHO criteria and motif-based method by a weighted average score, to benefit the advantages of different methods. Our results illustrated PREALW has better performance significantly in the crops' allergen prediction. This integrative allergen prediction algorithm could be useful for critical food safety matters. The PREALW could be accessed at http://lilab.life.sjtu.edu.cn:8080/prealw .

Navigating through the Jungle of Allergens: Features and Applications of Allergen Databases

The increasing number of available data on allergenic proteins demanded the establishment of structured, freely accessible allergen databases. In this review article, features and applications of 6 of the most widely used allergen databases are discussed. The WHO/IUIS Allergen Nomenclature Database is the official resource of allergen designations. Allergome is the most comprehensive collection of data on allergens and allergen sources. AllergenOnline is aimed at providing a peer-reviewed database of allergen sequences for prediction of allergenicity of proteins, such as those planned to be inserted into genetically modified crops. The Structural Database of Allergenic Proteins (SDAP) provides a database of allergen sequences, structures, and epitopes linked to bioinformatics tools for sequence analysis and comparison. The Immune Epitope Database (IEDB) is the largest repository of T-cell, B-cell, and major histocompatibility complex protein epitopes including epitopes of allergens. AllFam classifies allergens into families of evolutionarily related proteins using definitions from the Pfam protein family database. These databases contain mostly overlapping data, but also show differences in terms of their targeted users, the criteria for including allergens, data shown for each allergen, and the availability of bioinformatics tools.

An Overview of Bioinformatics Tools and Resources in Allergy

The rapidly increasing number of characterized allergens has created huge demands for advanced information storage, retrieval, and analysis. Bioinformatics and machine learning approaches provide useful tools for the study of allergens and epitopes prediction, which greatly complement traditional laboratory techniques. The specific applications mainly include identification of B- and T-cell epitopes, and assessment of allergenicity and cross-reactivity. In order to facilitate the work of clinical and basic researchers who are not familiar with bioinformatics, we review in this chapter the most important databases, bioinformatic tools, and methods with relevance to the study of allergens.

A novel lipid transfer protein from the pea Pisum sativum: isolation, recombinant expression, solution structure, antifungal activity, lipid binding, and allergenic properties

BACKGROUND

Plant lipid transfer proteins (LTPs) assemble a family of small (7-9 kDa) ubiquitous cationic proteins with an ability to bind and transport lipids as well as participate in various physiological processes including defense against phytopathogens. They also form one of the most clinically relevant classes of plant allergens. Nothing is known to date about correlation between lipid-binding and IgE-binding properties of LTPs. The garden pea Pisum sativum is widely consumed crop and important allergenic specie of the legume family. This work is aimed at isolation of a novel LTP from pea seeds and characterization of its structural, functional, and allergenic properties.

RESULTS

Three novel lipid transfer proteins, designated as Ps-LTP1-3, were found in the garden pea Pisum sativum, their cDNA sequences were determined, and mRNA expression levels of all the three proteins were measured at different pea organs. Ps-LTP1 was isolated for the first time from the pea seeds, and its complete amino acid sequence was determined. The protein exhibits antifungal activity and is a membrane-active compound that causes a leakage from artificial liposomes. The protein binds various lipids including bioactive jasmonic acid. Spatial structure of the recombinant uniformly (13)C,(15)N-labelled Ps-LTP1 was solved by heteronuclear NMR spectroscopy. In solution the unliganded protein represents the mixture of two conformers (relative populations ~ 85:15) which are interconnected by exchange process with characteristic time ~ 100 ms. Hydrophobic residues of major conformer form a relatively large internal tunnel-like lipid-binding cavity (van der Waals volume comes up to ~1000 Å(3)). The minor conformer probably corresponds to the protein with the partially collapsed internal cavity.

CONCLUSIONS

For the first time conformational heterogeneity in solution was shown for an unliganded plant lipid transfer protein. Heat denaturation profile and simulated gastrointestinal digestion assay showed that Ps-LTP1 displayed a high thermal and digestive proteolytic resistance proper for food allergens. The reported structural and immunological findings seem to describe Ps-LTP1 as potential cross-reactive allergen in LTP-sensitized patients, mostly Pru p 3(+) ones. Similarly to allergenic LTPs the potential IgE-binding epitope of Ps-LTP1 is located near the proposed entrance into internal cavity and could be involved in lipid-binding.

In silico allergenicity prediction of several lipid transfer proteins

Non-specific lipid transfer proteins (nsLTPs) are common allergens and they are particularly widespread within the plant kingdom. They have a highly conserved three-dimensional structure that generate a strong cross-reactivity among the members of this family. In the last years several web tools for the prediction of allergenicity of new molecules based on their homology with known allergens have been released, and guidelines to assess potential allergenicity of proteins through bioinformatics have been established. Even if such tools are only partially reliable yet, they can provide important indications when other kinds of molecular characterization are lacking. The potential allergenicity of 28 amino acid sequences of LTPs homologs, either retrieved from the UniProt database or in silico deduced from the corresponding EST coding sequence, was predicted using 7 publicly available web tools. Moreover, their similarity degree to their closest known LTP allergens was calculated, in order to evaluate their potential cross-reactivity. Finally, all sequences were studied for their identity degree with the peach allergen Pru p 3, considering the regions involved in the formation of its known conformational IgE-binding epitope. Most of the analyzed sequences displayed a high probability to be allergenic according to all the software employed. The analyzed LTPs from bell pepper, cassava, mango, mungbean and soybean showed high homology (>70%) with some known allergenic LTPs, suggesting a potential risk of cross-reactivity for sensitized individuals. Other LTPs, like for example those from canola, cassava, mango, mungbean, papaya or persimmon, displayed a high degree of identity with Pru p 3 within the consensus sequence responsible for the formation, at three-dimensional level, of its major conformational epitope. Since recent studies highlighted how in patients mono-sensitized to peach LTP the levels of IgE seem directly proportional to the chance of developing cross-reactivity to LTPs from non-Rosaceae foods, and these chances increase the more similar the protein is to Pru p 3, these proteins should be taken into special account for future studies aimed at evaluating the risk of cross-allergenicity in highly sensitized individuals.

1:1 FASTA update: Using the power of E-values in FASTA to detect potential allergen cross-reactivity

In the context of regulatory assessment of transgenic proteins for potential allergenicity, a previous investigation demonstrated that a 1:1 FASTA comparison using an E-value of 1.0E-09 as a criterion is superior to the conventional FASTA search (using the whole sequence as a query) for >35% identity over 80 amino acids, but with improved specificity. A further study, using groups of known cross-reactive peanut allergens, indicates the sensitivity of this approach is superior to the conventional FASTA search and equivalent to 80-mer sliding window FASTA search recommended by WHO/FAO. Specifically, the 1:1 FASTA approach eliminated the technical issues resulting from lack of identification of short query sequences with high identity to known allergens, or high identity over short amino acid stretches, and different E-value settings when searching for >35% identity over 80aa. Based on the performance of this simple application of existing bioinformatics tools, and its ease of implementation and interpretation in the context of a regulatory assessment, we advocate that adoption of this 1:1 FASTA approach as a supplement to the FAO/WHO/ CODEX criterion (>35% identity over 80aa) formulated 13 years ago. Adoption of this approach eliminates many biologically irrelevant homology hits generated by the FAO/WHO/CODEX criterion and improves the safety assessment of GM crops.

Purification, characterization and safety assessment of the introduced cold shock protein B in DroughtGard maize

DroughtGard maize was developed through constitutive expression of cold shock protein B (CSPB) from Bacillus subtilis to improve performance of maize (Zea mays) under water-limited conditions. B. subtilis commonly occurs in fermented foods and CSPB has a history of safe use. Safety studies were performed to further evaluate safety of CSPB introduced into maize. CSPB was compared to proteins found in current allergen and protein toxin databases and there are no sequence similarities between CSPB and known allergens or toxins. In order to validate the use of Escherichia coli-derived CSPB in other safety studies, physicochemical and functional characterization confirmed that the CSPB produced by DroughtGard possesses comparable molecular weight, immunoreactivity, and functional activity to CSPB produced from E. coli and that neither is glycosylated. CSPB was completely digested with sequential exposure to pepsin and pancreatin for 2 min and 30 s, respectively, suggesting that CSPB will be degraded in the mammalian digestive tract and would not be expected to be allergenic. Mice orally dosed with CSPB at 2160 mg/kg, followed by analysis of body weight gains, food consumption and clinical observations, showed no discernible adverse effects. This comprehensive safety assessment indicated that the CSPB protein from DroughtGard is safe for food and feed consumption.

Allergenic characterization of a novel allergen, homologous to chymotrypsin, from german cockroach

Purpose

Cockroach feces are known to be rich in IgE-reactive components. Various protease allergens were identified by proteomic analysis of German cockroach fecal extract in a previous study. In this study, we characterized a novel allergen, a chymotrypsin-like serine protease.

Methods

A cDNA sequence homologous to chymotrypsin was obtained by analysis of German cockroach expressed sequence tag (EST) clones. The recombinant chymotrypsins from the German cockroach and house dust mite (Der f 6) were expressed in Escherichia coli using the pEXP5NT/TOPO vector system, and their allergenicity was investigated by ELISA.

Results

The deduced amino acid sequence of German cockroach chymotrypsin showed 32.7 to 43.1% identity with mite group 3 (trypsin) and group 6 (chymotrypsin) allergens. Sera from 8 of 28 German cockroach allergy subjects (28.6%) showed IgE binding to the recombinant protein. IgE binding to the recombinant cockroach chymotrypsin was inhibited by house dust mite chymotrypsin Der f 6, while it minimally inhibited the German cockroach whole body extract.

Conclusions

A novel allergen homologous to chymotrypsin was identified from the German cockroach and was cross-reactive with Der f 6.

AllerTOP v.2--a server for in silico prediction of allergens

Allergy is an overreaction by the immune system to a previously encountered, ordinarily harmless substance--typically proteins--resulting in skin rash, swelling of mucous membranes, sneezing or wheezing, or other abnormal conditions. The use of modified proteins is increasingly widespread: their presence in food, commercial products, such as washing powder, and medical therapeutics and diagnostics, makes predicting and identifying potential allergens a crucial societal issue. The prediction of allergens has been explored widely using bioinformatics, with many tools being developed in the last decade; many of these are freely available online. Here, we report a set of novel models for allergen prediction utilizing amino acid E-descriptors, auto- and cross-covariance transformation, and several machine learning methods for classification, including logistic regression (LR), decision tree (DT), naïve Bayes (NB), random forest (RF), multilayer perceptron (MLP) and k nearest neighbours (kNN). The best performing method was kNN with 85.3% accuracy at 5-fold cross-validation. The resulting model has been implemented in a revised version of the AllerTOP server (http://www.ddg-pharmfac.net/AllerTOP).

Reduction of the number of major representative allergens: from clinical testing to 3-dimensional structures

Vast amounts of allergen sequence data have been accumulated, thus complicating the identification of specific allergenic proteins when performing diagnostic allergy tests and immunotherapy. This study aims to rank the importance/potency of the allergens so as to logically reduce the number of allergens and/or allergenic sources. Meta-analysis of 62 allergenic sources used for intradermal testing on 3,335 allergic patients demonstrated that in southern China, mite, sesame, spiny amaranth, Pseudomonas aeruginosa, and house dust account for 88.0% to 100% of the observed positive reactions to the 62 types of allergenic sources tested. The Kolmogorov-Smironov Test results of the website-obtained allergen data and allergen family featured peptides suggested that allergen research in laboratories worldwide has been conducted in parallel on many of the same species. The major allergens were reduced to 21 representative allergens, which were further divided into seven structural classes, each of which contains similar structural components. This study therefore has condensed numerous allergenic sources and major allergens into fewer major representative ones, thus allowing for the use of a smaller number of allergens when conducting comprehensive allergen testing and immunotherapy treatments.

Fungi: the neglected allergenic sources

Allergic diseases are considered the epidemics of the twentieth century estimated to affect more than 30% of the population in industrialized countries with a still increasing incidence. During the past two decades, the application of molecular biology allowed cloning, production and characterization of hundreds of recombinant allergens. In turn, knowledge about molecular, chemical and biologically relevant allergens contributed to increase our understanding of the mechanisms underlying IgE-mediated type I hypersensitivity reactions. It has been largely demonstrated that fungi are potent sources of allergenic molecules covering a vast variety of molecular structures including enzymes, toxins, cell wall components and phylogenetically highly conserved cross-reactive proteins. Despite the large knowledge accumulated and the compelling evidence for an involvement of fungal allergens in the pathophysiology of allergic diseases, fungi as a prominent source of allergens are still largely neglected in basic research as well as in clinical practice. This review aims to highlight the impact of fungal allergens with focus on asthma and atopic dermatitis.

Characterization and phylogenetic analysis of allergenic Tryp_alpha_amyl protein family in plants

Most known allergenic proteins in rice ( Oryza sativa ) seed belong to the Tryp_alpha_amyl family (PF00234), but the sequence characterization and the evolution of the allergenic Tryp_alpha_amyl family members in plants have not been fully investigated. In this study, two specific motifs were found besides the common alpha-amylase inhibitors (AAI) domain from the allergenic Tryp_alpha_amyl family members in rice seeds (trRSAs). To understand the evolution and functional importance of the Tryp_alpha_amy1 family and the specific motifs for the allergenic one, a BLAST search identified 75 homologous proteins of trRSAs (trHAs) from 22 plant species including main crops such as rice, maize ( Zea mays ), wheat ( Triticum aestivum ), and sorghum ( Sorghum bicolor ) from all available sequences in the public databases. Statistical analysis showed that the allergenicity of trHAs is closely associated with these two motifs with high number of cysteine residues (p value = 0.00026), and the trHAs with and without the two motifs were clustered into separate clades, respectively. Furthermore, significant difference was observed on the secondary and tertiary structures of allergenic and nonallergenic trHAs. In addition, expression analysis showed that trHA-encoding genes of purple false brome ( Brachypodium distachyon ), barrel medic ( Medicago truncatula ), rice, and sorghum are dominantly expressed in seeds. This work provides insight into the understanding of the properties of allergens in the Tryp_alpha_amyl family and is helpful for allergy therapy.

Allerdictor: fast allergen prediction using text classification techniques

MOTIVATION

Accurately identifying and eliminating allergens from biotechnology-derived products are important for human health. From a biomedical research perspective, it is also important to identify allergens in sequenced genomes. Many allergen prediction tools have been developed during the past years. Although these tools have achieved certain levels of specificity, when applied to large-scale allergen discovery (e.g. at a whole-genome scale), they still yield many false positives and thus low precision (even at low recall) due to the extreme skewness of the data (allergens are rare). Moreover, the most accurate tools are relatively slow because they use protein sequence alignment to build feature vectors for allergen classifiers. Additionally, only web server implementations of the current allergen prediction tools are publicly available and are without the capability of large batch submission. These weaknesses make large-scale allergen discovery ineffective and inefficient in the public domain.

RESULTS

We developed Allerdictor, a fast and accurate sequence-based allergen prediction tool that models protein sequences as text documents and uses support vector machine in text classification for allergen prediction. Test results on multiple highly skewed datasets demonstrated that Allerdictor predicted allergens with high precision over high recall at fast speed. For example, Allerdictor only took ∼6 min on a single core PC to scan a whole Swiss-Prot database of ∼540 000 sequences and identified <1% of them as allergens.

AVAILABILITY AND IMPLEMENTATION

Allerdictor is implemented in Python and available as standalone and web server versions at http://allerdictor.vbi.vt.edu CONTACT: lawrence@vbi.vt.edu Supplementary information: Supplementary data are available at Bioinformatics online.

In silico prediction of allergenic proteins

Currently, the prediction of new allergens is becoming important due to use of genetically modified (GM) foods and biopharmaceuticals. In this chapter, we describe how to use four popular allergenic prediction servers: (1) Structural Database of Allergenic Proteins (SDAP), (2) Allermatch, (3) Evaller 2, and (4) AlgPred. The first two prediction servers are based on traditional approaches, whereas Evaller 2 and AlgPred use sophisticated machine learning techniques.

Immunoinformatics: a brief review

A large volume of data relevant to immunology research has accumulated due to sequencing of genomes of the human and other model organisms. At the same time, huge amounts of clinical and epidemiologic data are being deposited in various scientific literature and clinical records. This accumulation of the information is like a goldmine for researchers looking for mechanisms of immune function and disease pathogenesis. Thus the need to handle this rapidly growing immunological resource has given rise to the field known as immunoinformatics. Immunoinformatics, otherwise known as computational immunology, is the interface between computer science and experimental immunology. It represents the use of computational methods and resources for the understanding of immunological information. It not only helps in dealing with huge amount of data but also plays a great role in defining new hypotheses related to immune responses. This chapter reviews classical immunology, different databases, and prediction tool. Further, it briefly describes applications of immunoinformatics in reverse vaccinology, immune system modeling, and cancer diagnosis and therapy. It also explores the idea of integrating immunoinformatics with systems biology for the development of personalized medicine. All these efforts save time and cost to a great extent.

PREAL: prediction of allergenic protein by maximum Relevance Minimum Redundancy (mRMR) feature selection

BACKGROUND

Assessment of potential allergenicity of protein is necessary whenever transgenic proteins are introduced into the food chain. Bioinformatics approaches in allergen prediction have evolved appreciably in recent years to increase sophistication and performance. However, what are the critical features for protein's allergenicity have been not fully investigated yet.

RESULTS

We presented a more comprehensive model in 128 features space for allergenic proteins prediction by integrating various properties of proteins, such as biochemical and physicochemical properties, sequential features and subcellular locations. The overall accuracy in the cross-validation reached 93.42% to 100% with our new method. Maximum Relevance Minimum Redundancy (mRMR) method and Incremental Feature Selection (IFS) procedure were applied to obtain which features are essential for allergenicity. Results of the performance comparisons showed the superior of our method to the existing methods used widely. More importantly, it was observed that the features of subcellular locations and amino acid composition played major roles in determining the allergenicity of proteins, particularly extracellular/cell surface and vacuole of the subcellular locations for wheat and soybean. To facilitate the allergen prediction, we implemented our computational method in a web application, which can be available at http://gmobl.sjtu.edu.cn/PREAL/index.php.

CONCLUSIONS

Our new approach could improve the accuracy of allergen prediction. And the findings may provide novel insights for the mechanism of allergies.

AllergenFP: allergenicity prediction by descriptor fingerprints

MOTIVATION

Allergenicity, like antigenicity and immunogenicity, is a property encoded linearly and non-linearly, and therefore the alignment-based approaches are not able to identify this property unambiguously. A novel alignment-free descriptor-based fingerprint approach is presented here and applied to identify allergens and non-allergens. The approach was implemented into a four step algorithm. Initially, the protein sequences are described by amino acid principal properties as hydrophobicity, size, relative abundance, helix and β-strand forming propensities. Then, the generated strings of different length are converted into vectors with equal length by auto- and cross-covariance (ACC). The vectors were transformed into binary fingerprints and compared in terms of Tanimoto coefficient.

RESULTS

The approach was applied to a set of 2427 known allergens and 2427 non-allergens and identified correctly 88% of them with Matthews correlation coefficient of 0.759. The descriptor fingerprint approach presented here is universal. It could be applied for any classification problem in computational biology. The set of E-descriptors is able to capture the main structural and physicochemical properties of amino acids building the proteins. The ACC transformation overcomes the main problem in the alignment-based comparative studies arising from the different length of the aligned protein sequences. The conversion of protein ACC values into binary descriptor fingerprints allows similarity search and classification.

AVAILABILITY AND IMPLEMENTATION

The algorithm described in the present study was implemented in a specially designed Web site, named AllergenFP (FP stands for FingerPrint). AllergenFP is written in Python, with GIU in HTML. It is freely accessible at http://ddg-pharmfac.net/Allergen FP.

CONTACT

idoytchinova@pharmfac.net or ivanbangov@shu-bg.net.

AllerTOP--a server for in silico prediction of allergens

Background

Allergy is a form of hypersensitivity to normally innocuous substances, such as dust, pollen, foods or drugs. Allergens are small antigens that commonly provoke an IgE antibody response. There are two types of bioinformatics-based allergen prediction. The first approach follows FAO/WHO Codex alimentarius guidelines and searches for sequence similarity. The second approach is based on identifying conserved allergenicity-related linear motifs. Both approaches assume that allergenicity is a linearly coded property. In the present study, we applied ACC pre-processing to sets of known allergens, developing alignment-independent models for allergen recognition based on the main chemical properties of amino acid sequences.

Results

A set of 684 food, 1,156 inhalant and 555 toxin allergens was collected from several databases. A set of non-allergens from the same species were selected to mirror the allergen set. The amino acids in the protein sequences were described by three z-descriptors (z₁, z₂ and z₃) and by auto- and cross-covariance (ACC) transformation were converted into uniform vectors. Each protein was presented as a vector of 45 variables. Five machine learning methods for classification were applied in the study to derive models for allergen prediction. The methods were: discriminant analysis by partial least squares (DA-PLS), logistic regression (LR), decision tree (DT), naïve Bayes (NB) and k nearest neighbours (kNN). The best performing model was derived by kNN at k = 3. It was optimized, cross-validated and implemented in a server named AllerTOP, freely accessible at http://www.pharmfac.net/allertop. AllerTOP also predicts the most probable route of exposure. In comparison to other servers for allergen prediction, AllerTOP outperforms them with 94% sensitivity.

Conclusions

AllerTOP is the first alignment-free server for in silico prediction of allergens based on the main physicochemical properties of proteins. Significantly, as well allergenicity AllerTOP is able to predict the route of allergen exposure: food, inhalant or toxin.

Evaluation and integration of existing methods for computational prediction of allergens

Background

Allergy involves a series of complex reactions and factors that contribute to the development of the disease and triggering of the symptoms, including rhinitis, asthma, atopic eczema, skin sensitivity, even acute and fatal anaphylactic shock. Prediction and evaluation of the potential allergenicity is of importance for safety evaluation of foods and other environment factors. Although several computational approaches for assessing the potential allergenicity of proteins have been developed, their performance and relative merits and shortcomings have not been compared systematically.

Results

To evaluate and improve the existing methods for allergen prediction, we collected an up-to-date definitive dataset consisting of 989 known allergens and massive putative non-allergens. The three most widely used allergen computational prediction approaches including sequence-, motif- and SVM-based (Support Vector Machine) methods were systematically compared using the defined parameters and we found that SVM-based method outperformed the other two methods with higher accuracy and specificity. The sequence-based method with the criteria defined by FAO/WHO (FAO: Food and Agriculture Organization of the United Nations; WHO: World Health Organization) has higher sensitivity of over 98%, but having a low specificity. The advantage of motif-based method is the ability to visualize the key motif within the allergen. Notably, the performances of the sequence-based method defined by FAO/WHO and motif eliciting strategy could be improved by the optimization of parameters. To facilitate the allergen prediction, we integrated these three methods in a web-based application proAP, which provides the global search of the known allergens and a powerful tool for allergen predication. Flexible parameter setting and batch prediction were also implemented. The proAP can be accessed at http://gmobl.sjtu.edu.cn/proAP/main.html.

Conclusions

This study comprehensively evaluated sequence-, motif- and SVM-based computational prediction approaches for allergens and optimized their parameters to obtain better performance. These findings may provide helpful guidance for the researchers in allergen-prediction. Furthermore, we integrated these methods into a web application proAP, greatly facilitating users to do customizable allergen search and prediction.