Nature: Over 1,700 New Microproteins Critical for Cancer Survival Discovered in 'Junk' DNA
A study conducted by a consortium of leading global institutions (including the Princess Máxima Center and the University of Michigan) has revealed that a quarter of non-coding DNA regions produce 'microproteins.' Silencing the genetic instructions for six of these microproteins led to the death of 85% of cancer cells, opening up an entirely new generation of targets for immunotherapy and cancer vaccines.
The Dark Proteome: Why 1,700 New Microproteins Will Rewrite the Rules in Oncology and Beyond
The Bottom Line: What's Really Happening
On May 9, 2026, a paper was published in Nature by the TransCODE consortium—an international group of over 60 researchers from more than 30 institutions, including the Princess Máxima Center (Netherlands), the University of Michigan, the European Bioinformatics Institute (EMBL-EBI), and MIT. They sequenced 7,264 non-canonical open reading frames (ncORFs) in so-called 'junk' DNA and found that roughly a quarter of them—over 1,700—actually produce protein molecules.
This is not an abstract finding. The scientists tested the functionality of six such microproteins via CRISPR screening. The result: disabling one of them (OLMALINC) caused 85% of the 485 tested cancer cell lines to lose their ability to survive. Most of the discovered peptideins are displayed on the cell surface, making them ready-made targets for immunotherapy.
The consortium introduced a new term, 'peptidein,' for these molecules—an intermediate class between short peptides and full-fledged proteins. All data have been made publicly available in the GENCODE, UniProt, and PeptideAtlas databases.
Timeline and Context
The story of the 'dark proteome' didn't start yesterday. In 2024, M. Mar Albà's group from Barcelona published a study in Science Advances showing that non-canonical ORFs in hepatocellular carcinoma produce microproteins that become tumor-specific antigens in over 40% of patients. Around the same time, it was found that lncRNA-derived peptides exhibit tumor selectivity and immunogenicity at the level of CD8+ T-cell response.
Concurrently, in 2025, a group from NYU presented the ImmunoVerse atlas—28,446 tumor-specific HLA-presented antigens, of which 5,928 were previously unannotated and originated from 'dark' genomic regions. In March 2026, Frontiers in RNA Research published a review describing uORF-encoded peptides as a new class of neoantigens for personalized vaccines.
The current publication in Nature is the culmination of this trend. TransCODE didn't just find a few more microproteins. The consortium systematically mapped the entire pool of ncORFs, conducted functional screens, and created a reference database for the entire scientific community. This is a shift from 'blind spots' to systematic knowledge.
The key genetic locus is OLMALINC, which was previously thought to be a non-coding RNA. It is now known to encode a peptidein critical for cell division and DNA damage response. Interestingly, its role in normal cells remains unclear—this opens a window for selective targeting of tumors without systemic toxicity.
At the same time, researchers from the Princess Máxima Center had previously identified another peptidein playing a critical role in medulloblastoma—an aggressive pediatric brain tumor. This means the findings apply not only to adult oncology but also to pediatric cases.
Who Wins and Who Loses
Winners:
- Biotech companies working in neoantigen vaccines. This involves dozens of startups. Leaders in the field include BioNTech (market cap around $28 billion) and Moderna (around $35 billion). Both companies have invested in personalized neoantigen vaccine platforms. Peptideins expand the pool of targets manifold—especially for tumors with low mutational burden, where classical neoantigens are simply insufficient for vaccine development.
- Immunopeptidomics platforms. Companies like OHMX.bio, which perform multi-omics analysis to discover 'dark antigens,' receive scientific validation of their business approach. Their market niche—integrating genomics, transcriptomics, ribosome profiling, and immunopeptidomics—is becoming mainstream.
- Owners of CRISPR platforms for functional screens. Editas Medicine, Intellia Therapeutics, CRISPR Therapeutics—all need new targets. 1,700 new candidates, six of which have already shown functional significance, provide a roadmap for the next five years. The market for CRISPR-based screening services is estimated at around $2.7 billion by 2028.
Losers:
- Classical proteome databases. UniProt, PeptideAtlas, and GENCODE must now incorporate peptideins as a separate record class. This requires revising annotation algorithms and mass spectrometry search pipelines. Traditional pipelines tuned only to canonical ORFs have missed these molecules for years. Now, anyone using standard databases without the 'dark' extension works with inherently incomplete data.
- Investors who bet on 'classical' neoantigen approaches. Mutational neoantigens (missense mutations) have been mainstream for the past five years. But the Barcelona group's work in 2024 showed that mutational antigens in hepatocellular carcinoma are 99.4% private—unique to each patient. Peptideins from lncRNA exhibit 28.4% patient-shared antigens. This means 'dark' antigens are better suited for off-the-shelf vaccines, while mutational ones are only for personalized ones.
- Pharma companies with narrow pipelines. Those who invested exclusively in canonical protein targets risk being left with an outdated portfolio. Peptideins represent thousands of new targets, each requiring separate preclinical programs. Companies without R&D flexibility will lose the race to those who adapt quickly.
What the Media Isn't Saying
The first non-obvious insight concerns the term 'peptidein.' Most media outlets present it as a curiosity—'scientists invented a new word.' In reality, it's a strategic move. Creating a formal taxonomic category forces regulators—FDA and EMA—to acknowledge the existence of this class of molecules. And regulatory recognition opens the door to accelerated approval procedures (breakthrough therapy designation, orphan drug status). The TransCODE consortium isn't just publishing science; it's laying the legal and regulatory foundation for future applications. Notably, among the authors are bioinformatics specialists from EMBL-EBI—an organization whose databases are reference standards for European regulators.
The second unspoken point: the CRISPR screen showed that disabling OLMALINC kills 85% of cancer cells. But no one in the press release mentions what happens in normal cells. The phrase 'its role in normal, healthy cells remains unclear' is a euphemism. It means the consortium either didn't test normal tissues or got ambiguous results they don't want to disclose until further studies. If OLMALINC turns out to be necessary for bone marrow stem cells, targeting it could be risky—hematological toxicity might outweigh the benefits.
The third point: the data are publicly available. This is a noble gesture, but also a way to 'stake a claim.' The consortium has effectively published the coordinates of 1,700 new targets. Those who start patent searches on them first will gain an advantage. Expect a wave of patent filings from biotech companies in the coming weeks.
The fourth insight concerns the Princess Máxima Center's previous work on medulloblastoma. Pediatric oncology is a small market niche but huge in reputation. It's the easiest path to orphan drug designation and accelerated approval. Likely, the first clinical applications of peptideins will be in pediatrics—not because the need is greater there, but because the regulatory pathway is faster.
Forecast: Next 30 Days and 90 Days
30 days (by June 9, 2026):
- EMBL-EBI will complete the integration of peptideins into GENCODE and UniProt as a separate category. This is a purely technical step, but it will unlock automatic detection of peptideins by standard bioinformatics pipelines. Currently, they are only found by specialized methods.
- At least three major cancer centers (MD Anderson, Memorial Sloan Kettering, Gustave Roussy) will announce screening programs for peptideins in their tumor sample collections. Each has biobanks of 50,000+ samples. This will create an avalanche of data.
- The first patent application for a therapeutic antibody against one of the six functionally validated peptideins will be filed—most likely through MIT or the University of Michigan, whose researchers participated in the consortium.
90 days (by August 7, 2026):
- BioNTech or Moderna will announce the inclusion of peptideins in the design of their next-generation neoantigen vaccine. For BioNTech, this is a logical step after the success of FixVac—their vaccine based on shared tumor antigens. Peptideins expand their target library by an order of magnitude. Expected announcement in a quarterly report.
- Consolidation of startups in the 'dark proteome' field will begin. OHMX.bio, Tesorai, and several European biotechs may become acquisition targets for major players—Illumina, Thermo Fisher, Bio-Rad. Estimated deal values range from $120–400 million.
- The FDA will receive the first Investigational New Drug (IND) application for a therapeutic agent targeting a peptidein. Given the orphan drug direction, the first target will likely be medulloblastoma or another pediatric tumor with high unmet need.
- One of the six validated peptideins will undergo toxicity validation in normal tissues. If even one shows no cross-reactivity with vital organs, it will catalyze $200–300 million in venture capital funding for the field as a whole.
Fundamental takeaway: we are witnessing not just the 'discovery of new proteins.' We are present at a structural shift in oncology—from the 'one target, one drug' model to a 'hidden target atlas' model. The dark proteome is a reservoir that doubles or triples the number of potential therapeutic intervention points. In 90 days, this will cease to be an academic sensation and will begin to become an industrial reality.
— Editorial Team
