15.06.2026
The field of precision oncology has reached a critical tipping point. While the FDA approval of drugs targeting MHC-presented peptides has firmly validated this class of therapeutics, a persistent barrier remains: patient eligibility. Because traditional Human Leukocyte Antigen (HLA) alleles vary drastically across global populations, many groundbreaking immunotherapies only work for a fraction of patients.
Alithea Bio is thrilled to have contributed to a major breakthrough that directly addresses this challenge. In collaboration with Binbin and the visionary team at Vcreate, a new preprint introduces MHC Attention—a deep learning and high-throughput screening tool designed to unlock the universal potential of HLA-E cancer immunotherapy.
Why HLA-E is a Game-Changer for Cancer Vaccines and TCR Drugs
Unlike highly polymorphic classical HLA alleles, HLA-E is remarkably conserved. In fact, the specific HLA-E*01 variant is carried by roughly 99% of the human population.
This near-universal coverage means that a therapeutic peptide presented by HLA-E could theoretically treat almost any patient, eliminating the narrow demographic constraints of first-generation MHC drugs.
Historically, mapping these targets has been incredibly difficult due to sparse training data and technical limitations in identifying which peptides are successfully presented by HLA-E. The collaborative development of MHC Attention changes everything.
Our Contribution: Fueling AI with Real-World Human Tissue Data
An AI model is only as good as the data used to train it. While computational insights are valuable, neural networks require deep, high-quality biological realities to learn effectively.
Alithea Bio’s Core Contribution: We provided the critical foundation for this breakthrough—our proprietary immunopeptidomics data generated directly from real human tissue samples.
By feeding MHC Attention with our highly accurate, real-world tissue data rather than synthetic or simulated inputs, the neural network was able to learn true biological rules of presentation. This real-world human data bridge is what allowed the platform to successfully move from predictive theory to validated clinical targets.
Inside the Technology: MHC Attention and High-Throughput Screening
To bridge the data gap and identify novel targets, the collaborative research team combined cutting-edge machine learning with advanced functional assays:
MHC Attention Architecture: A specialized neural network that learns allele-level attention directly from primary tissue data, significantly increasing prediction accuracy.
Pooled Mammalian Cell Assays: An innovative screening method capable of testing thousands of HLA-E peptides within a single, streamlined experiment.
Validated Discoveries: The pipeline has already successfully identified and validated highly promising, novel HLA-E cancer peptides derived from well-known oncogenic drivers, including WT1, ETV4, RNF43, and BMP8A.
Key Takeaways from the Preprint
Universal Applications: Moves the industry away from niche, allele-specific treatments toward off-the-shelf therapies.
Validated Classes: Leverages an FDA-approved therapeutic mechanism but expands its addressable market to 99% of patients.
Supported Research: This pivotal work was backed by prominent institutions, including the National Science Foundation (NSF), National Cancer Institute (NCI), National Institute of General Medical Sciences (NIGMS), and the Merck Digital Sciences Studio.
Leverage Alithea Bio’s Proprietary Human Tissue Database
At Alithea Bio, our mission is to turn real-world biological complexity into validated clinical assets. Led by our executive team—CEO Fanny Giannou, CSO Dr. Tim Fugmann, and CFO Hagen Grauer—we specialize in providing the deep mass spectrometry-based immunopeptidomics data required to clear the bottleneck of target discovery.
Whether you are targeting classical alleles or exploring the massive potential of HLA-E, our proprietary human tissue database can validate your therapeutic candidates against genuine healthy and tumor profiles with next-day turnaround times.
“Unlocking conserved alleles like HLA-E requires shifting away from purely synthetic models and anchoring our AI in true human tissue data.” — Dr. Tim Fugmann, CSO.
Explore the Research
The tool and findings are now openly accessible to the scientific community:
Read the Preprint: Dive deep into the full scientific methodology and data in the official preprint: MHC Attention: Identifying HLA-E–presented cancer antigens through deep learning and high-throughput screening.
Try the Tool: Experience the web platform firsthand at the Vcreate MHC Attention Suite.
Want to find out how Alithea Bio’s proprietary human tissue immunopeptidomics platform can support your HLA-E or TCR discovery pipeline? Contact our team today to schedule a technical consultation.
