A team of researchers spent years watching their quantum circuits fail before one finally worked. In early 2025, scientists from the Wellcome Sanger Institute, the University of Oxford, the University of Cambridge, the University of Melbourne, and Kyiv Academic University successfully encoded the complete genome of the Hepatitis D virus onto IBM’s 156-qubit Heron quantum processor. The Sanger Institute called it a world first: a natural genome, loaded in its entirety onto a quantum computer.
The virus genome is small, just roughly 1,700 nucleotide bases. A human genome contains more than three billion. But the method the team developed was built to bridge that gap, and the researchers say they have already shown, on paper, how it could scale to the size of a full human genome. Whether it actually will is the question driving a 12-month sprint that began when the project was selected for the final phase of Wellcome Leap’s Q4Bio Challenge.
Years of dead ends
The path to this result was not smooth. Earlier attempts to use quantum computing for genomics ran into brutal scaling problems. A 2021 peer-reviewed study in Scientific Reports documented how quantum annealing and quantum-inspired methods hit hard walls when genome sizes increased. Circuit complexity ballooned, resource demands became impractical, and the approaches that worked on synthetic or tiny data sets collapsed under anything resembling real biological sequences.
Those failures were not wasted. They mapped out exactly where the bottlenecks were, and the Oxford-led team designed its new approach as a direct response. The technique, based on a mathematical framework called Matrix Product State (MPS) formalism, generates quantum circuits that scale with genome length in a controlled way rather than demanding exponentially more resources as sequences grow longer. It exploits patterns and redundancy in DNA data to compress genetic information into a form that fits on a fixed number of qubits.
A preprint paper on arXiv by Creevey, Hassan, McCafferty, Hollenberg, and Strelchuk lays out the method in detail. As a proof of concept, the authors demonstrate how the bacteriophage PhiX174 genome, which has 5,386 nucleotide bases, can be compressed into a compact quantum description. The key selling point: the circuit-generation method does not require a complete redesign for each new genome. In principle, it can handle sequences of vastly different sizes using the same underlying architecture.
What the Hepatitis D experiment actually showed
The Hepatitis D encoding was the first time this MPS-based approach was pushed onto a large, state-of-the-art quantum processor. IBM’s 156-qubit Heron chip is among the company’s most advanced devices, and running the encoding on real hardware rather than a simulator exposed the method to the noise and imperfections that plague every current quantum machine.
That distinction matters. Quantum processors at this scale are notoriously error-prone. Every operation introduces small inaccuracies, and those errors compound across a circuit. The Sanger Institute’s announcement confirms the encoding was completed, but no publicly available data from the run specifies how accurately the processor represented the genome. Without published error metrics, fidelity benchmarks, or details on how many repetitions were needed to achieve a stable result, the practical quality of the encoding cannot be independently assessed.
This is not unusual for an early-stage demonstration, but it is a gap that outside researchers will want closed before drawing strong conclusions.
The enormous distance still to cover
Even with a working method, the scale challenge is staggering. The Hepatitis D genome is roughly 1,700 bases. The PhiX174 bacteriophage is about 5,386. A single human chromosome can contain hundreds of millions of base pairs, and the full human genome tops three billion. Moving from a viral genome to a human one is not a matter of incremental improvement. It requires the MPS compression to hold up across orders of magnitude more data, on hardware that does not yet exist at the necessary scale and reliability.
Classical computers, by comparison, already handle human genomes routinely. Modern sequencing pipelines can align, assemble, and analyze a full human genome in hours. The case for quantum genomics is not that classical tools cannot do the job today, but that certain problems, such as comparing many genomes simultaneously in pangenome analysis, modeling complex structural variations, or searching vast genomic databases, may eventually benefit from quantum speedups that classical architectures cannot match.
A review published in April 2025 in npj Genomic Medicine examined this intersection carefully. The authors catalogued potential applications including faster variant detection, personalized treatment matching, and pangenome analysis, but emphasized that none have been demonstrated on any quantum platform. They highlighted limited device access, hardware-specific bottlenecks, and the wide gap between research-grade demonstrations and the reliability that clinical workflows demand.
Loading data is not the same as analyzing it
There is a distinction that often gets lost in coverage of quantum milestones: encoding data onto a quantum computer is not the same as doing something useful with it. The MPS method addresses what researchers call the state-preparation problem, getting biological data into a form that quantum circuits can process. That is a necessary first step, but it is only a first step.
The next challenge is running algorithms on that encoded data that extract meaningful biological insights. Tasks like variant calling (identifying mutations), haplotype phasing (determining which gene variants sit on the same chromosome), or detecting large structural rearrangements in DNA all require algorithmic designs that can exploit quantum advantages while tolerating the noise levels of current hardware. None of those algorithms have been demonstrated on quantum-encoded genomic data yet.
Sergii Strelchuk of the University of Oxford, who leads the Quantum Pangenomics project, now has a concrete deadline to push past this boundary. The Wellcome Leap Q4Bio final phase gives the team roughly 12 months to show that quantum-encoded genomes can yield results that biologists would recognize as useful analysis, not just successful data loading.
What the next 12 months need to deliver
For anyone tracking quantum biology, that deadline is the most important detail in this story. Within the next year, observers should expect clearer reporting on error rates from the Hepatitis D run, demonstrations of at least basic computational tasks performed on encoded genomes, and more detailed documentation of which institutions contributed what to the research pipeline. (Cornell University, for instance, appears in the citation trail for the MPS paper, but no primary source specifies what its researchers contributed.)
If those results materialize, the Hepatitis D milestone will look like the foundation of a genuinely new computational toolkit for genomics. If they do not, it will remain a technically impressive but narrow proof of concept, a reminder that quantum computing’s promise for biology is still measured more in potential than in practice.
The genome is loaded. Now the real work begins.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
