Next-generation IEDB tools: a platform for epitope prediction and analysis.

The Next-Generation (NG) IEDB Tools website (https://nextgen-tools.iedb.org) provides users with a redesigned interface to many of the algorithms for epitope prediction and analysis that were originally released on the legacy IEDB Tools website. The initial release focuses on consolidation of all tools related to HLA class I epitopes (MHC binding, elution, immunogenicity, and processing), making all of these predictions accessible from a single application and allowing for their simultaneous execution with minimal user inputs. Additionally, the PEPMatch tool for identifying highly similar epitopes in a set of curated proteomes, as well as a tool for epitope clustering, are available on the site. The NG Tools site allows users to build data pipelines by sending the output of one tool as input for the next. Over the next several years, all pre-existing IEDB Tools, and any newly developed tools, will be integrated into this new site. Here we describe the philosophy behind the redesign and demonstrate the utility and productivity enhancements that are enabled by the new interface.

Autoimmunity Against Surfactant Protein B Is Associated with Pneumonitis During Checkpoint Blockade.

RATIONALE: Immune checkpoint inhibitor-related pneumonitis is a serious autoimmune event affecting up to 20% of patients with non-small cell lung cancer, yet the factors underpinning its development in some patients and not others are poorly understood. OBJECTIVES: To investigate the role of autoantibodies and autoreactive T cells against surfactant-related proteins in the development of pneumonitis. METHODS: The study cohort consisted of non-small cell lung cancer patients who gave blood samples before and during immune checkpoint inhibitor treatment. Serum was used for proteomics analyses and to detect autoantibodies present during pneumonitis. T cell stimulation assays and single-cell RNA sequencing were performed to investigate the specificity and functionality of peripheral autoreactive T cells. The findings were confirmed in a validation cohort comprising patients with non-small cell lung cancer and patients with melanoma. MEASUREMENTS AND MAIN RESULTS: Across both cohorts, patients who developed pneumonitis had higher pre-treatment levels of immunoglobulin G autoantibodies targeting surfactant protein-B. At the onset of pneumonitis, these patients also exhibited higher frequencies of CD4+ interferon-gamma-positive surfactant protein B-specific T cells, and expanding T cell clonotypes recognizing this protein, accompanied by a pro-inflammatory serum proteomic profile. CONCLUSIONS: Our data suggest that the co-occurrence of surfactant protein-B-specific immunoglobulin G autoantibodies and CD4+ T cells is associated with the development of pneumonitis during ICI therapy. Pre-treatment levels of these antibodies may represent a potential biomarker for elevated risk of developing pneumonitis and on-treatment levels may provide a diagnostic aid. This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives License 4.0 (http://creativecommons.org/licenses/by-nc-nd/4.0/).

The Cancer Epitope Database and Analysis Resource (CEDAR).

We established The Cancer Epitope Database and Analysis Resource (CEDAR) to catalog all epitope data in the context of cancer. The specific molecular targets of adaptive T cell and B cell immune responses are referred to as epitopes. Epitopes derived from cancer antigens are of high relevance as they are recognized by anti-cancer immune cells. Detailed knowledge of the molecular characteristic of cancer epitopes and associated metadata is relevant to understanding and planning prophylactic and therapeutic applications and accurately characterizing naturally occurring immune responses and cancer immunopathology. CEDAR provides a freely accessible, comprehensive collection of cancer epitope and receptor data curated from the literature and serves as a companion site to the Immune Epitope Database (IEDB), which is focused on infectious, autoimmune, and allergic diseases. CEDAR is freely accessible at https://cedar.iedb.org/.

PEPMatch: a tool to identify short peptide sequence matches in large sets of proteins.

BACKGROUND: Numerous tools exist for biological sequence comparisons and search. One case of particular interest for immunologists is finding matches for linear peptide T cell epitopes, typically between 8 and 15 residues in length, in a large set of protein sequences. Both to find exact matches or matches that account for residue substitutions. The utility of such tools is critical in applications ranging from identifying conservation across viral epitopes, identifying putative epitope targets for allergens, and finding matches for cancer-associated neoepitopes to examine the role of tolerance in tumor recognition. RESULTS: We defined a set of benchmarks that reflect the different practical applications of short peptide sequence matching. We evaluated a suite of existing methods for speed and recall and developed a new tool, PEPMatch. The tool uses a deterministic k-mer mapping algorithm that preprocesses proteomes before searching, achieving a 50-fold increase in speed over methods such as the Basic Local Alignment Search Tool (BLAST) without compromising recall. PEPMatch's code and benchmark datasets are publicly available. CONCLUSIONS: PEPMatch offers significant speed and recall advantages for peptide sequence matching. While it is of immediate utility for immunologists, the developed benchmarking framework also provides a standard against which future tools can be evaluated for improvements. The tool is available at https://nextgen-tools.iedb.org , and the source code can be found at https://github.com/IEDB/PEPMatch .

Allele-Specific Thresholds of Eluted Ligands for T-Cell Epitope Prediction.

A common strategy for predicting candidate human leukocyte antigen class I T-cell epitopes is to use an affinity-based threshold of 500 nM. Although a 500 nM threshold across alleles effectively identifies most epitopes across alleles, findings showing that major histocompatibility complex repertoire sizes vary by allele indicate that using thresholds specific to individual alleles may improve epitope identification. In this work, we compare different strategies utilizing common and allele-specific thresholds to identify an optimal approach for T-cell epitope prediction. First, we confirmed previous observations that different human leukocyte antigen class I alleles correspond with varying repertoire sizes. Here, we define general and allele-specific thresholds that capture 80% of eluted ligands and a different set of thresholds associated with capturing 9-mer T-cell epitopes at 80% sensitivity. Our analysis revealed that allele-specific threshold performance was roughly equivalent to that of a common threshold when considering a relatively large number of alleles. However, when predicting epitopes for only a few alleles, the use of allele-specific thresholds would be preferable. Finally, we present here for public use a set of allele-specific thresholds that may be used to identify T-cell epitopes at 80% sensitivity.

Impact of Cysteine Residues on MHC Binding Predictions and Recognition by Tumor-Reactive T Cells.

The availability of MHC-binding prediction tools has been useful in guiding studies aimed at identifying candidate target Ags to generate reactive T cells and to characterize viral and tumor-reactive T cells. Nevertheless, prediction algorithms appear to function poorly for epitopes containing cysteine (Cys) residues, which can oxidize and form disulfide bonds with other Cys residues under oxidizing conditions, thus potentially interfering with their ability to bind to MHC molecules. Analysis of the results of HLA-A*02:01 class I binding assays carried out in the presence and absence of the reducing agent 2-ME indicated that the predicted affinity for 25% of Cys-containing epitopes was underestimated by a factor of 3 or more. Additional analyses were undertaken to evaluate the responses of human CD8+ tumor-reactive T cells against 10 Cys-containing HLA class I-restricted minimal determinants containing substitutions of α-aminobutyric acid (AABA), a cysteine analogue containing a methyl group in place of the sulfhydryl group present in Cys, for the native Cys residues. Substitutions of AABA for Cys at putative MHC anchor positions often significantly enhanced T cell recognition, whereas substitutions at non-MHC anchor positions were neutral, except for one epitope where this modification abolished T cell recognition. These findings demonstrate the need to evaluate MHC binding and T cell recognition of Cys-containing peptides under conditions that prevent Cys oxidation, and to adjust current prediction binding algorithms for HLA-A*02:01 and potentially additional class I alleles to more accurately rank peptides containing Cys anchor residues.

Comparison of HLA ligand elution data and binding predictions reveals varying prediction performance for the multiple motifs recognized by HLA-DQ2.5.

Binding prediction tools are commonly used to identify peptides presented on MHC class II molecules. Recently, a wealth of data in the form of naturally eluted ligands has become available and discrepancies between ligand elution data and binding predictions have been reported. Quantitative metrics for such comparisons are currently lacking. In this study, we assessed how efficiently MHC class II binding predictions can identify naturally eluted peptides, and investigated instances with discrepancies between the two methods in detail. We found that, in general, MHC class II eluted ligands are predicted to bind to their reported restriction element with high affinity. But, for several studies reporting an increased number of ligands that were not predicted to bind, we found that the reported MHC restriction was ambiguous. Additional analyses determined that most of the ligands predicted to not bind, are predicted to bind other co-expressed MHC class II molecules. For selected alleles, we addressed discrepancies between elution data and binding predictions by experimental measurements and found that predicted and measured affinities correlate well. For DQA1*05:01/DQB1*02:01 (DQ2.5) however, binding predictions did miss several peptides that were determined experimentally to be binders. For these peptides and several known DQ2.5 binders, we determined key residues for conferring DQ2.5 binding capacity, which revealed that DQ2.5 utilizes two different binding motifs, of which only one is predicted effectively. These findings have important implications for the interpretation of ligand elution data and for the improvement of MHC class II binding predictions.

In vitro induction of neoantigen-specific T cells in myelodysplastic syndrome, a disease with low mutational burden.

Therapies that utilize immune checkpoint inhibition work by leveraging mutation-derived neoantigens and have shown greater clinical efficacy in tumors with higher mutational burden. Whether tumors with a low mutational burden are susceptible to neoantigen-targeted therapy has not been fully addressed. To examine the feasibility of neoantigen-specific adoptive T-cell therapy, the authors studied the T-cell response against somatic variants in five patients with myelodysplastic syndrome (MDS), a malignancy with a very low tumor mutational burden. DNA and RNA from tumor (CD34+) and normal (CD3+) cells isolated from the patients' blood were sequenced to predict patient-specific MDS neopeptides. Neopeptides representing the somatic variants were used to induce and expand autologous T cells ex vivo, and these were systematically tested in killing assays to determine the proportion of neopeptides yielding neoantigen-specific T cells. The authors identified a total of 32 somatic variants (four to eight per patient) and found that 21 (66%) induced a peptide-specific T-cell response and 19 (59%) induced a T-cell response capable of killing autologous tumor cells. Of the 32 somatic variants, 11 (34%) induced a CD4+ response and 11 (34%) induced a CD8+ response that killed the tumor. These results indicate that in vitro induction of neoantigen-specific T cells is feasible for tumors with very low mutational burden and that this approach warrants investigation as a therapeutic option for such patients.

A large-scale study of peptide features defining immunogenicity of cancer neo-epitopes.

Accurate prediction of immunogenicity for neo-epitopes arising from a cancer associated mutation is a crucial step in many bioinformatics pipelines that predict outcome of checkpoint blockade treatments or that aim to design personalised cancer immunotherapies and vaccines. In this study, we performed a comprehensive analysis of peptide features relevant for prediction of immunogenicity using the Cancer Epitope Database and Analysis Resource (CEDAR), a curated database of cancer epitopes with experimentally validated immunogenicity annotations from peer-reviewed publications. The developed model, ICERFIRE (ICore-based Ensemble Random Forest for neo-epitope Immunogenicity pREdiction), extracts the predicted ICORE from the full neo-epitope as input, i.e. the nested peptide with the highest predicted major histocompatibility complex (MHC) binding potential combined with its predicted likelihood of antigen presentation (%Rank). Key additional features integrated into the model include assessment of the BLOSUM mutation score of the neo-epitope, and antigen expression levels of the wild-type counterpart which is often reflecting a neo-epitope's abundance. We demonstrate improved and robust performance of ICERFIRE over existing immunogenicity and epitope prediction models, both in cross-validation and on external validation datasets.

A functional identification platform reveals frequent, spontaneous neoantigen-specific T cell responses in patients with cancer.

The clinical impact of tumor-specific neoantigens as both immunotherapeutic targets and biomarkers has been impeded by the lack of efficient methods for their identification and validation from routine samples. We have developed a platform that combines bioinformatic analysis of tumor exomes and transcriptional data with functional testing of autologous peripheral blood mononuclear cells (PBMCs) to simultaneously identify and validate neoantigens recognized by naturally primed CD4+ and CD8+ T cell responses across a range of tumor types and mutational burdens. The method features a human leukocyte antigen (HLA)-agnostic bioinformatic algorithm that prioritizes mutations recognized by patient PBMCs at a greater than 40% positive predictive value followed by a short-term in vitro functional assay, which allows interrogation of 50 to 75 expressed mutations from a single 50-ml blood sample. Neoantigens validated by this method include both driver and passenger mutations, and this method identified neoantigens that would not have been otherwise detected using an in silico prediction approach. These findings reveal an efficient approach to systematically validate clinically actionable neoantigens and the T cell receptors that recognize them and demonstrate that patients across a variety of human cancers have a diverse repertoire of neoantigen-specific T cells.

Estimating tissue-specific peptide abundance from public RNA-Seq data.

Several novel MHC class I epitope prediction tools additionally incorporate the abundance levels of the peptides' source antigens and have shown improved performance for predicting immunogenicity. Such tools require the user to input the MHC alleles and peptide sequences of interest, as well as the abundance levels of the peptides' source proteins. However, such expression data is often not directly available to users, and retrieving the expression level of a peptide's source antigen from public databases is not trivial. We have developed the Peptide eXpression annotator (pepX), which takes a peptide as input, identifies from which proteins the peptide can be derived, and returns an estimate of the expression level of those source proteins from selected public databases. We have also investigated how the abundance level of a peptide can be best estimated in cases when it can originate from multiple transcripts and proteins and found that summing up transcript-level expression values performs best in distinguishing ligands from decoy peptides.

IEDB and CEDAR: Two Sibling Databases to Serve the Global Scientific Community.

Various methodologies have been utilized to analyze epitope-specific responses in the context of non-self-antigens, such as those associated with infectious diseases and allergies, and in the context of self-antigens, such as those associated with transplantation, autoimmunity, and cancer. Further to this, epitope-specific data, and its associated immunological context, are crucial to training and developing predictive algorithms and pipelines for the development of specific vaccines and diagnostics. In this chapter, we describe the methodology utilized to derive two sibling resources, the Immune Epitope Database (IEDB) and Cancer Epitope Database and Analysis Resource (CEDAR), to specifically host this data, and make them freely available to the scientific community.

A meta-analysis of epitopes in prostate-specific antigens identifies opportunities and knowledge gaps.

BACKGROUND: The Cancer Epitope Database and Analysis Resource (CEDAR) is a newly developed repository of cancer epitope data from peer-reviewed publications, which includes epitope-specific T cell, antibody, and MHC ligand assays. Here we focus on prostate cancer as our first cancer category to demonstrate the capabilities of CEDAR, and to shed light on the advances of epitope-related prostate cancer research. RESULTS: The meta-analysis focused on a subset of data describing epitopes from 8 prostate-specific (PS) antigens. A total of 460 epitopes were associated with these proteins, 187 T cell, 109B cell, and 271 MHC ligand epitopes. The number of epitopes was not correlated with the length of the protein; however, we found a significant positive correlation between the number of references per specific PS antigen and the number of reported epitopes. Forty-four different class I and 27 class II restrictions were found, with the most epitopes described for HLA-A*02:01 and HLA-DRB1*01:01. Cytokine assays were mostly limited to IFNg assays and a very limited number of tetramer assays were performed. Monoclonal and polyclonal B cell responses were balanced, with the highest number of epitopes studied in ELISA/Western blot assays. Additionally, epitopes were generically described as associated with prostate cancer, with little granularity specifying diseases state. We found that in vivo and tumor recognition assays were sparse, and the number of epitopes with annotated B/T cell receptor information were limited. Potential immunodominant regions were identified by the use of the ImmunomeBrowser tool. CONCLUSION: CEDAR provides a comprehensive repository of epitopes related to prostate-specific antigens. This inventory of epitope data with its wealth of searchable T cell, B cell and MHC ligand information provides a useful tool for the scientific community. At the same time, we identify significant knowledge gaps that could be addressed by experimental analysis.

The role of antigen expression in shaping the repertoire of HLA presented ligands.

Human leukocyte antigen (HLA) presentation of peptides is a prerequisite of T cell immune activation. The understanding of the rules defining this event has large implications for our knowledge of basic immunology and for the rational design of immuno-therapeutics and vaccines. Historically, most of the available prediction methods have been solely focused on the information related to antigen processing and presentation. Recent work has, however, demonstrated that method performance can be boosted by integrating information related to antigen abundance. Here we expand on these later findings and develop an extended version of NetMHCpan, called NetMHCpanExp, integrating information on antigen abundance from RNA-Seq experiments. In line with earlier works, the model demonstrates improved performance for both HLA ligand and cancer neoantigen epitope prediction. Optimal results are obtained by use of sample-specific abundance information but also reference datasets can be applied with a limited performance drop. The developed tool is available at https://services.healthtech.dtu.dk/service.php?NetMHCpanExp-1.0.

Combined assessment of MHC binding and antigen abundance improves T cell epitope predictions.

Many steps of the MHC class I antigen processing pathway can be predicted using computational methods. Here we show that epitope predictions can be further improved by considering abundance levels of peptides' source proteins. We utilized biophysical principles and existing MHC binding prediction tools in concert with abundance estimates of source proteins to derive a function that estimates the likelihood of a peptide to be an MHC class I ligand. We found that this combination improved predictions for both naturally eluted ligands and cancer neoantigen epitopes. We compared the use of different measures of antigen abundance, including mRNA expression by RNA-Seq, gene translation by Ribo-Seq, and protein abundance by proteomics on a dataset of SARS-CoV-2 epitopes. Epitope predictions were improved above binding predictions alone in all cases and gave the highest performance when using proteomic data. Our results highlight the value of incorporating antigen abundance levels to improve epitope predictions.

Predicting the Success of Fmoc-Based Peptide Synthesis.

Synthetic peptides are commonly used in biomedical science for many applications in basic and translational research. While peptide synthesis is generally easy and reliable, the chemical nature of some amino acids as well as the many steps and chemical compounds involved can render the synthesis of some peptide sequences difficult. Identification of these problematic sequences and mitigation of issues they may present can be important for the reliable use of peptide reagents in several contexts. Here, we assembled a large dataset of peptides that were synthesized using standard Fmoc chemistry and whose identity was validated using mass spectrometry. We analyzed the mass spectra to identify errors in peptide syntheses and sought to develop a computational tool to predict the likelihood that any given peptide sequence would be synthesized accurately. Our model, named Peptide Synthesis Score (PepSySco), is able to predict the likelihood that a peptide will be successfully synthesized based on its amino acid sequence.

TCRMatch: Predicting T-Cell Receptor Specificity Based on Sequence Similarity to Previously Characterized Receptors.

The adaptive immune system in vertebrates has evolved to recognize non-self antigens, such as proteins expressed by infectious agents and mutated cancer cells. T cells play an important role in antigen recognition by expressing a diverse repertoire of antigen-specific receptors, which bind epitopes to mount targeted immune responses. Recent advances in high-throughput sequencing have enabled the routine generation of T-cell receptor (TCR) repertoire data. Identifying the specific epitopes targeted by different TCRs in these data would be valuable. To accomplish that, we took advantage of the ever-increasing number of TCRs with known epitope specificity curated in the Immune Epitope Database (IEDB) since 2004. We compared seven metrics of sequence similarity to determine their power to predict if two TCRs have the same epitope specificity. We found that a comprehensive k-mer matching approach produced the best results, which we have implemented into TCRMatch, an openly accessible tool (http://tools.iedb.org/tcrmatch/) that takes TCR ß-chain CDR3 sequences as an input, identifies TCRs with a match in the IEDB, and reports the specificity of each match. We anticipate that this tool will provide new insights into T cell responses captured in receptor repertoire and single cell sequencing experiments and will facilitate the development of new strategies for monitoring and treatment of infectious, allergic, and autoimmune diseases, as well as cancer.

The Cancer Epitope Database and Analysis Resource: A Blueprint for the Establishment of a New Bioinformatics Resource for Use by the Cancer Immunology Community.

Recent years have witnessed a dramatic rise in interest towards cancer epitopes in general and particularly neoepitopes, antigens that are encoded by somatic mutations that arise as a consequence of tumorigenesis. There is also an interest in the specific T cell and B cell receptors recognizing these epitopes, as they have therapeutic applications. They can also aid in basic studies to infer the specificity of T cells or B cells characterized in bulk and single-cell sequencing data. The resurgence of interest in T cell and B cell epitopes emphasizes the need to catalog all cancer epitope-related data linked to the biological, immunological, and clinical contexts, and most importantly, making this information freely available to the scientific community in a user-friendly format. In parallel, there is also a need to develop resources for epitope prediction and analysis tools that provide researchers access to predictive strategies and provide objective evaluations of their performance. For example, such tools should enable researchers to identify epitopes that can be effectively used for immunotherapy or in defining biomarkers to predict the outcome of checkpoint blockade therapies. We present here a detailed vision, blueprint, and work plan for the development of a new resource, the Cancer Epitope Database and Analysis Resource (CEDAR). CEDAR will provide a freely accessible, comprehensive collection of cancer epitope and receptor data curated from the literature and provide easily accessible epitope and T cell/B cell target prediction and analysis tools. The curated cancer epitope data will provide a transparent benchmark dataset that can be used to assess how well prediction tools perform and to develop new prediction tools relevant to the cancer research community.

De novo mutations in mitochondrial DNA of iPSCs produce immunogenic neoepitopes in mice and humans.

The utility of autologous induced pluripotent stem cell (iPSC) therapies for tissue regeneration depends on reliable production of immunologically silent functional iPSC derivatives. However, rejection of autologous iPSC-derived cells has been reported, although the mechanism underlying rejection is largely unknown. We hypothesized that de novo mutations in mitochondrial DNA (mtDNA), which has far less reliable repair mechanisms than chromosomal DNA, might produce neoantigens capable of eliciting immune recognition and rejection. Here we present evidence in mice and humans that nonsynonymous mtDNA mutations can arise and become enriched during reprogramming to the iPSC stage, long-term culture and differentiation into target cells. These mtDNA mutations encode neoantigens that provoke an immune response that is highly specific and dependent on the host major histocompatibility complex genotype. Our results reveal that autologous iPSCs and their derivatives are not inherently immunologically inert for autologous transplantation and suggest that iPSC-derived products should be screened for mtDNA mutations.

Predicting T cell recognition of MHC class I restricted neoepitopes.

Epitopes that arise from a somatic mutation, also called neoepitopes, are now known to play a key role in cancer immunology and immunotherapy. Recent advances in high-throughput sequencing have made it possible to identify all mutations and thereby all potential neoepitope candidates in an individual cancer. However, most of these neoepitope candidates are not recognized by T cells of cancer patients when tested in vivo or in vitro, meaning they are not immunogenic. Especially in patients with a high mutational load, usually hundreds of potential neoepitopes are detected, highlighting the need to further narrow down this candidate list. In our study, we assembled a dataset of known, naturally processed, immunogenic neoepitopes to dissect the properties that make these neoepitopes immunogenic. The tools to use and thresholds to apply for prioritizing neoepitopes have so far been largely based on experience with epitope identification in other settings such as infectious disease and allergy. Here, we performed a detailed analysis on our dataset of curated immunogenic neoepitopes to establish the appropriate tools and thresholds in the cancer setting. To this end, we evaluated different predictors for parameters that play a role in a neoepitope's immunogenicity and suggest that using binding predictions and length-rescaling yields the best performance in discriminating immunogenic neoepitopes from a background set of mutated peptides. We furthermore show that almost all neoepitopes had strong predicted binding affinities (as expected), but more surprisingly, the corresponding non-mutated peptides had nearly as high affinities. Our results provide a rational basis for parameters in neoepitope filtering approaches that are being commonly used. Abbreviations: SNV: single nucleotide variant; nsSNV: nonsynonymous single nucleotide variant; ROC: receiver operating characteristic; AUC: area under ROC curve; HLA: human leukocyte antigen; MHC: major histocompatibility complex; PD-1: Programmed cell death protein 1; PD-L1 or CTLA-4: cytotoxic T-lymphocyte associated protein 4.