Dark antigens: Expanding the horizons of immunopeptidomics and multi-omics discovery

Introduction to dark antigens

The antigenic universe accessible to T cells extends far beyond the conventional proteome. Dark antigens, also referred to as cryptic, hidden, or noncanonical antigens, are by-products of genomic and translational events outside conventional annotated protein-coding space and therefore occult to traditional neoantigen pipelines. Because they are outside conventional discovery pathways, these cryptic antigens represent a vast and largely untapped reservoir of immunogenic targets. Data indicates that most cancer-restricted epitopes are found in the “dark” antigen space, offering new therapeutic opportunities. Mass spectrometry analysis reveals cryptic peptides make up the majority of the immunopeptidome, with dominant immunogenic ones in certain tumors. The cryptic immunopeptidome is increasingly seen as a major contributor to tumor specificity, with many cryptic MAPs (MHC-associated peptides) arising from splicing abnormalities, noncoding RNAs, untranslated regions, and epigenetically deregulated loci^1-7. OHMX.bio aims to illuminate this frontier by integrating immunopeptidomics with genomics, transcriptomics, epigenomics, and ribosome profiling, thereby enabling systematic discovery of dark antigens and accelerating their translation into immunotherapy.

What are dark antigens?

Dark antigens represent unconventional translation products that are presented on major histocompatibility complex class I (MHC-I) molecules. These peptides arise when eukaryotic ribosomes translate genomic regions beyond annotated open reading frames (ORFs), generating novel antigens not captured by standard proteomic analyses. Non-canonical MHC-associated peptides (ncMAPs) are produced from aberrantly translated small ORFs (sORFs) in non-protein-coding regions like untranslated regions, long non-coding RNAs, and pseudogenes. These dark antigens extend beyond conventional proteomics, generating peptides from introns, reading frames, pseudogenes, noncoding RNA transcripts, and non-AUG initiation codons. Most tumour-specific MHC-associated peptides in this dark genomic compartment are derived from genomic loci with limited expression in normal tissues, causing strong central tolerance and strong Cytotoxic T Lymphocyte (CTL) responses in most cancers.

These cryptic proteins are encoded by short ORFs, which frequently start translation from non-AUG codons that are nearly identical. The biological relevance of cryptic proteins is demonstrated by the fact that they are more frequently detected in the immunopeptidome than their canonical counterparts. Strong options for the development of universal or semi-universal cancer vaccines are presented by the non-canonical peptides that are shared by a variety of malignancies. Experimental studies reveal tumor immunopeptidomes, including pancreatic cancer, contain dark or cryptic peptides that induce direct tumor cell killing through CTL activation. Vaccination with these peptides has shown robust anti-tumour effects in preclinical models, revealing numerous nonmutated peptides from previously non-coding genomic regions. These dark antigens exhibit tumour-specific expression and recurrent detection, making them promising candidates for next-generation cancer immunotherapy^1-4.

The role of multi-omics in discovering dark antigens

Discovering dark antigens is inherently more complex than conventional neoantigens, because their sources lie outside standard annotation. Mass‐spectrometry immunopeptidomics alone can miss these signals but combining it with transcriptomics and genomics dramatically expands discovery. Layered omics has revealed peptides from cryptic open reading frames in untranslated regions, introns, noncoding RNAs and alternative splice isoforms. Dark peptides, often elusive in database searches due to noncanonical sources like alternative reading frames or untranslated regions, are identified when sample-specific transcriptomes are translated and combined with genomic data variant calls. This immunopeptidogenomic strategy has demonstrated the ability to recover the vast majority of known HLA ligands while uncovering hundreds of previously undetected peptides. A hidden layer of immunological targets beyond the standard proteome can be revealed by researchers mapping the whole landscape of presenting antigens by combining immunopeptidomic data with RNA-Seq and DNA-Seq. This holistic approach is critical for developing precision immunotherapy and understanding tumor immune evasion mechanisms.

Bioinformatic pipelines are key to this multi-omics strategy. For example, a workflow of “immunopeptidogenomics” has been generated that employs RNA-Seq to construct a user-specific protein database by collapsing transcripts, reading them in three frames, and adding sample-specific mutations via variant calling. The resulting custom FASTA is then used to search the HLA‐peptide mass spectra. This pipeline relies on tools like STAR and GATK for alignment and mutation calling, and on in-house scripts to three-frame-translate transcripts and merge them into “cryptic” protein entries. In acute myeloid leukemia and similar contexts, advanced proteogenomic approaches have successfully recapitulated most standard peptides while identifying thousands of novel epitopes. End-to-end frameworks now integrate mass spectrometry-based immunopeptidomics with tumor DNA and RNA sequencing to predict and rank tumor-specific antigens, including canonical neoantigens, viral peptides, and noncanonical tumor-specific antigens, aiding vaccine development. By combining genomics, transcriptomics, and proteomics, these pipelines can detect peptides that would otherwise remain hidden.

Immunopeptidogenomics. Source: Katherine E. Scull, Kirti Pandey, Sri H. Ramarathinam, Anthony W. Purcell, Immunopeptidogenomics: Harnessing RNA-Seq to Illuminate the Dark Immunopeptidome, Molecular & Cellular Proteomics, Volume 20, 2021, 100143, ISSN 1535-9476, https://doi.org/10.1016/j.mcpro.2021.100143.

Researchers are using specific workflows to identify new HLA-I peptides originating from circular RNAs in various cancer types. These peptides screen for back-splice junctions, are stored in custom databases, and undergo rigorous mass spectrometry searches. The methods also involve translating entire transcriptomes, including the noncoding RNAs, and sorting the resulting mass spectrometry data against universal protein databases to eliminate known sequences. Integrating immunopeptidomics with genomics, transcriptomics, epigenomics, and translatomics is increasingly crucial to the identification of antigenic peptides which are typically ignored by conventional proteomic pipelines. This multi-omic approach enhances the resolution of the antigen landscape, helping to reveal new targets for cancer immunotherapy and advance personalized medicine approaches^7-9.

Applications in cancer and immunotherapy

Peptides from cancer cells outside of the canonical set are a new area of potential for immunotherapy. An ever-growing body of data suggests that antigen-presenting cells, including dendritic cells, present these “dark” peptides on MHC molecules and thus activate tumor-reactive T cells; along with this, prototype vaccines based on such peptides have been developed. In a striking proof-of-concept, T cells that were programmed to target cryptic peptides, which are translated from genomic regions that are classically regarded as noncoding, were found to slow tumor growth in pancreatic organoid systems. Analogously, mouse immunization with combinatorial pools of combinatorial tumor-restricted cryptic peptides induced intense CD8⁺ T-cell responses and caused significant tumor regression in preclinical models. These findings confirm that non-mutated dark antigens are indeed true immunogens: they are specifically expressed on tumor cells and, when presented to the immune system, elicit robust responses.

Cryptocytic peptides are guiding new diagnostics and treatments beyond vaccinations. Noncanonical peptides have been included in CAR-T designs to increase tumour selectivity and broaden TIL therapy. Methodical mapping of patient HLA ligand repertoires is identifying shared cryptic epitopes, suggesting potential diagnostic biomarkers or targets. High-resolution identification is being defined using deep sequencing, proteogenomic processes, and ribosome profiling (Ribo-seq). Coupled with mass spectrometry and de novo peptide discovery, comprehensive immunopeptidome atlases reveal thousands of noncanonical MHC class I–associated peptides (ncMAPs), derived from untranslated regions, long noncoding RNAs, upstream open reading frames, pseudogenes, introns, and alternative reading frames. Many of these peptides recur across tumor types and exhibit favorable attributes, high predicted MHC-binding affinity, optimal T-cell contact residues, and negligible expression in normal tissues. Modern computational frameworks enable weighted prioritization of candidate antigens, balancing tumor specificity, immunogenic potential, and safety constraints, demonstrating substantial tumor suppression in lung cancer preclinical models. In acute myeloid leukemia models, integrating both canonical and noncanonical antigen repertoires expands the therapeutic target space, with several noncanonical peptides ranking among the most predictive of immune infiltration and treatment response. Recent advancements have thus, reveal an expanding repertoire of epitopes, many of which are being translated into immunotherapeutic targets, providing a rich reservoir of antigens for cancer vaccines, engineered T cells, immune monitoring, and diagnostic use^5,6,10.

OHMX.bio’s role in dark antigen discovery

At OHMX.bio, our vision is to provide a state-of-the-art multi-omics-driven dark antigen discovery engine. Starting from tumor and matched normal samples, we can perform deep genomic profiling (WGS/WES) and transcriptomics (RNA-seq with splice and circular RNA resolution). We add ribosome profiling to capture translational evidence and epigenomics (ATAC-seq, methylation) to define regulatory context. We then construct a bespoke expanded reference library combining canonical and cryptic ORFs. Using advanced immunopeptidomics, we isolate MHC-bound peptides and search via motif-guided, inclusion-list–augmented pipelines (e.g. inspired by NeoDiscMS and IntroSpect). We apply machine learning and filtering (expression, translation confidence, binding affinity, immunogenicity predictors) to rank candidates. Finally, OHMX.bio validates top peptides via synthetic peptide binding assays, cellular T cell activation assays, and TCR screening. By closing the loop from DNA to peptide to immunological function, OHMX.bio positions itself as a pioneering partner in translating dark antigens into cancer vaccine and T cell therapy leads.

Conclusion

Dark antigens represent a transformative expansion of the antigenic universe, transforming occult genomic and translational events into actionable immunotherapeutic targets. By combining immunopeptidomics with genomics, transcriptomics, epigenomics, and ribosome profiling, it is possible to systematically identify cryptic epitopes that traditional methods miss. Dark antigens hold significant potential to improve vaccine coverage, refine TCR-based therapies, and uncover novel biomarkers, especially in cancers with low mutational burden. With its integrated multi-omics discovery platform and strong translational validation capabilities, OHMX.bio is well positioned to lead the next wave of immunotherapy innovation by illuminating the dark antigen frontier.

Frequently asked questions about dark antigens

What are dark antigens?

Dark antigens are noncanonical peptides that originate from genomic regions outside traditional protein-coding sequences. They may arise from untranslated regions, introns, alternative reading frames, pseudogenes, or long non-coding RNAs. Because conventional proteomics pipelines focus on annotated proteins, these peptides often remain undetected. However, many dark antigens are presented on MHC molecules and can trigger immune responses, making them promising targets in cancer immunotherapy.

Why is a multi-omics approach necessary to discover dark antigens?

Mass spectrometry alone may miss cryptic peptides. Therefore, dark antigen discovery requires integration of immunopeptidomics with genomics, transcriptomics, ribosome profiling, and epigenomics. By constructing sample-specific protein databases derived from RNA sequencing and variant calling, researchers can detect peptides that would otherwise remain invisible. This layered multi-omics strategy dramatically expands the antigen landscape and enables systematic discovery of novel immunogenic targets.

How can dark antigens improve cancer immunotherapy?

Dark antigens can serve as targets for personalized cancer vaccines, adoptive T-cell therapies, and engineered CAR-T or TCR therapies. Because these peptides often originate from regions with minimal expression in healthy tissue, they may bypass central immune tolerance and generate strong cytotoxic T-cell responses. Preclinical studies demonstrate that targeting cryptic peptides can reduce tumor growth and improve immune activation.

How does OHMX.bio identify dark antigens?

OHMX.bio combines deep genomic sequencing (WGS/WES), RNA sequencing, ribosome profiling, and epigenetic profiling with high-resolution immunopeptidomics. Expanded protein reference libraries include both canonical and cryptic open reading frames. Machine learning and motif-guided search pipelines rank candidate peptides, followed by functional validation using binding assays and T-cell activation studies. This end-to-end workflow enables translation of dark antigens into therapeutic strategies.

Are dark antigens relevant beyond oncology?

Yes. While particularly impactful in cancer immunotherapy, dark antigens may also be relevant in infectious diseases and autoimmune research. Expanding the detectable antigen repertoire improves understanding of immune recognition and may uncover novel diagnostic or therapeutic targets in multiple disease areas.

References

[1] Starck, S. R., & Shastri, N. (2016). Nowhere to hide: unconventional translation yields cryptic peptides for immune surveillance. Immunological reviews, 272(1), 8-16.

[2] Chong, C., Coukos, G., & Bassani-Sternberg, M. (2022). Identification of tumor antigens with immunopeptidomics. Nature biotechnology, 40(2), 175-188.

[3] Kina, E., Larouche, J. D., Thibault, P., & Perreault, C. (2025). The cryptic immunopeptidome in health and disease. Trends in Genetics.

[4] Saffern, M., & Samstein, R. (2023). MHCing the tumour’s dark genome. Nature reviews. Immunology, 23(3), 140.

[5] Raja, R., Mangalaparthi, K. K., Madugundu, A. K., Jessen, E., Pathangey, L., Magtibay, P., … & Curtis, M. (2025). Immunogenic cryptic peptides dominate the antigenic landscape of ovarian cancer. Science Advances, 11(8), eads7405.

[6] Ruzzi, F., Riccardo, F., Conti, L., Tarone, L., Semprini, M. S., Bolli, E., … & Cavallo, F. (2025). Cancer vaccines: Target antigens, vaccine platforms and preclinical models. Molecular Aspects of Medicine, 101, 101324.

[7] Huber, F., Arnaud, M., Stevenson, B. J., Michaux, J., Benedetti, F., Thevenet, J., … & Bassani-Sternberg, M. (2024). A comprehensive proteogenomic pipeline for neoantigen discovery to advance personalized cancer immunotherapy. Nature biotechnology, 1-13.

[8] Ferreira, H. J., Stevenson, B. J., Pak, H., Yu, F., Almeida Oliveira, J., Huber, F., … & Bassani-Sternberg, M. (2024). Immunopeptidomics-based identification of naturally presented non-canonical circRNA-derived peptides. Nature communications, 15(1), 2357.

[9] Scull, K. E., Pandey, K., Ramarathinam, S. H., & Purcell, A. W. (2021). Immunopeptidogenomics: Harnessing RNA-seq to illuminate the dark immunopeptidome. Molecular & Cellular Proteomics, 20, 100143.

[10] Schaeffeler, E., Walz, J., & Schwab, M. (2025). Personalized cancer vaccination is emerging: lessons learnt from renal cancer and challenges for broader application. Signal Transduction and Targeted Therapy, 10(1), 107.