Researchers at the University of Oxford have helped to achieve a world first: loading a complete genome onto a quantum computer. This makes an important step forward towards a future where quantum computing accelerates biological discovery.
Researchers have long sought to harness the potential of quantum computing for scientific discovery. Unlike classical computers, which process information as binary bits, quantum computers use qubits that can exist in multiple states simultaneously, allowing certain types of calculations to be explored far more efficiently.
However, current quantum systems remain highly sensitive and error-prone. Qubits are easily disrupted by noise and interference, making reliable large-scale computation extremely difficult. A major challenge for the field is demonstrating that today’s quantum hardware can still produce meaningful results for real-world scientific problems despite these limitations.
Genomics is one area where researchers believe quantum computing could eventually prove valuable. Modern genomic research increasingly relies on analysing vast collections of genetic data to understand disease pathways, population-level variation and biological processes. These datasets can become extraordinarily large and computationally demanding.
To explore how quantum computers might eventually address challenges in genomics and public health more broadly, Wellcome Leap launched the Quantum for Bio (Q4Bio) programme in 2023.
This 30-month, $50 million international challenge programme was designed to test whether useful biological and healthcare algorithms can be realised on current and emerging quantum hardware. Six teams reached the final phase, including the Quantum Pangenomics team led by Associate Professor Sergii Strelchuk at Oxford’s Department of Computer Science.
A pangenome is a collection of genome sequences from many individuals of the same species. Rather than representing a single reference genome, pangenomes capture the genetic diversity across many populations, providing a more complete view of genetic variation.
However, analysing multiple genomes at once dramatically increases computational complexity: as more genomes are incorporated into a pangenome, the burden on classical computational tools grows rapidly.
The goal of the Quantum Pangenomics project was to demonstrate the feasibility of using quantum computers to assemble genomes and pangenomes from DNA sequence data, as well as mapping DNA fragments into reference genomes in order to assess genetic variation.
The program was a collaboration between the Universities of Oxford, Cambridge and Melbourne, the Wellcome Sanger Institute and Kyiv Academic University. Together, they achieved a world first in loading the genome of the hepatitis D virus onto a quantum computer for the first time.
“Our goal has always been to push the boundaries of what’s possible in genomics. When we work with pangenomes, the information is presented in a form of a tangled maze, but we are building quantum algorithms to help find the best path through this maze when regular tools, such as classic computers, just get hopelessly stuck. So, we’re aiming for a simple but game-changing idea by bringing quantum computing into the world of genomics,” said Sergii Strelchuk, associate professor at the Department of Computer Science, University of Oxford.
The team chose the genome of the hepatitis D virus because it is highly compact (around 1,700 bases of RNA), while still representing a complete real-world genome. It therefore offered a manageable starting point for developing and testing methods to compress genomic data into quantum states. Hepatitis D is also clinically relevant, responsible for a severe liver infection that can be fatal.
Successfully loading the data onto IBM’s 156-qubit Heron processor required the researchers to create new approaches for representing genomic information within the constraints of current quantum hardware.
Unlike classical bits, qubits can exist in superpositions and can become entangled, meaning the behaviour of a multi-qubit system depends on correlations across many qubits rather than on each qubit separately.
This gives quantum computers a very large space in which to carry out computations. However, quantum computers do not simply act as exponentially larger memory devices: the information must be encoded in a form that can be prepared, manipulated and meaningfully measured.
For genomic data, the key challenge was therefore not only to compress the sequence, but to encode it in a way that preserves biologically relevant structure for future quantum algorithms.
In practice, however, loading genomic data onto a quantum computer was far from straightforward. Rather than simply transferring a sequence of DNA letters into the system, the genome first had to be converted into a suitable structure that could be represented on quantum hardware.
The team then had to design the precise sequence of operations needed to prepare this state on a real quantum processor. Ultimately, they succeeded in encoding the hepatitis D genome using 117 qubits.
“This was not just a data-loading exercise. The real challenge was to turn a biological sequence into quantum instructions that today’s hardware could actually run. That is what makes the result significant: it moves quantum genomics from a conceptual possibility towards an executable workflow,” said Strelchuk.
Looking ahead, the researchers believe the platform they have developed could help tackle some of the most computationally challenging problems in human health, including metagenomics and antimicrobial resistance.
More rapid and more powerful genomic analysis could help enable deeper understanding of rare genetic disorders, facilitate rapid tracking of infectious disease and allow precise identification of disease-causing genetic mutations.
In particular, the research team hope these approaches could contribute to understanding chromothripsis – a cancer mechanism in which chromosomes are shattered and incorrectly reassembled. This is an area of significant complexity that classical computational methods have so far struggled to address.
