DTU Scientists Generate Novel Peptides With Quantum-AI, Lifting Accuracy on Sparse Data
Updated
Updated · WIRED · Jul 12
DTU Scientists Generate Novel Peptides With Quantum-AI, Lifting Accuracy on Sparse Data
2 articles · Updated · WIRED · Jul 12
Summary
Laboratory tests showed the DTU team’s hybrid quantum-AI model produced more successful protein-binding peptides than a classical version, especially on targets with limited training data.
The researchers paired a generative protein model with ORCA Computing’s printer-sized quantum computer to widen peptide diversity—a key step toward vaccines and personalized immunotherapies.
Weekend work and leftover project funds drove the study, which its lead scientist said was needed to prove to skeptics that quantum predictions could translate into real biological results.
Quantum hardware still cannot run full-scale cutting-edge AI models, so the method is not yet transformative; the team will next test larger proteins and more advanced models.
The approach could matter most for understudied populations and neglected diseases, where scarce genetic data and limited funding make conventional drug-discovery models less effective.
With quantum-designed vaccines now a reality, how will this technology change our response to the next global pandemic?
As quantum AI designs personalized drugs, will it cure rare diseases or create a new 'quantum divide' in healthcare?
A desktop-sized quantum computer just outperformed classical AI. Is this the moment quantum computing finally leaves the lab?
InstaNovo+ Achieves 90% Accuracy in De Novo Peptide Sequencing: AI and Quantum-AI Set New Benchmarks for Proteomics and Beyond
Overview
In March 2025, scientists from the Technical University of Denmark and InstaDeep introduced InstaNovo and InstaNovo+, marking a major shift in peptide sequencing. Unlike traditional methods that rely on matching mass spectrometry data to known databases, these AI-powered models can directly reconstruct peptide sequences from raw data. This breakthrough allows researchers to discover new and modified peptides that were previously hidden, expanding our understanding of proteins and the unexplored 'dark matter' of the proteome. By overcoming the limits of older techniques, InstaNovo and InstaNovo+ open new possibilities for biological research and medical discovery.