AI Designed Peptides Emerge From Hybrid Quantum-Classical Design

Researchers from eight institutions across Denmark, the UK, and Poland have combined quantum computing with classical methods to design peptides capable of binding to major histocompatibility complex (MHC) class I molecules, a critical step in initiating immune responses. The team reports demonstrating a hybrid approach to de novo peptide design, addressing a key challenge in both vaccine development and T-cell therapies where identifying effective peptide candidates remains difficult, particularly for less-studied alleles. This work focuses on complex sequence spaces, exemplified by the HLA-A*31:01 allele linked to severe drug reactions, where limited training data hinders traditional computational methods. By leveraging photonic quantum processors and sampling from high-dimensional distributions, the researchers aim to broaden…