Finding Quantum Computing Applications in Technology Development

Quantum computers can be many times more powerful than classical computers, but which real-world applications they excel at often isn’t obvious. We’ve used one type of quantum computing to guide the development of algorithms in spatial biology. That’s one “real-world application” of quantum computing – what’s yours, and how can we help with it?

Quantum computers use common sense-defying phenomena arising on the quantum scale of atoms and sub-atomic particles to perform operations.

In classical computing, the fundamental unit of information, a bit, can only exist in one state at a time (0 or 1), while a quantum bit or qubit can exist in a superposition of states (0, 1, or a combination of the two). Along with the other peculiar abilities derived from quantum phenomena, this gives quantum computers certain advantages over classical computers.

In particular, quantum computers have the potential to resolve problems which require examination of many, many possibilities in a fraction of the time it would take the most powerful classical computers to do so.

For example, if a classical computer is tasked with finding the shortest path connecting a number of given locations, it would approach the problem by examining each possible path in turn. With each branching possibility – in this case, perhaps a choice to proceed to one location instead of another – the number of possibilities doubles: the computational effort scales exponentially!

A quantum computer, however, would approach the problem by examining every possible path at once. In theory, quantum computers can solve combinatorial optimisation problems which classical computers would not be able to solve in the lifetime of the universe.

‘What are quantum computers good for?’ is, according to a 2022 McKinsey report, a 700-billion-dollar question, and engineers all over the world are trying to work out how best to utilise the quantum advantage. Despite these efforts, the impact on real-life applications is still limited.

At TTP, we recognise the benefits of this ground breaking technology and we have put it to good use to guide the development of classical algorithms for spatial biology.

Quantum computers that operate on quantum-bits with logical gates – similar to what classical computers do on bits – are called “universal quantum computers”. They offer the most generic platform to perform quantum operations. However, this comes at a cost. Driving such a system is complex, and they are sensitive to external noise.

Quantum annealers, in contrast, are less flexible but more stable – instability of qubits being a major practical barrier to quantum computing – and are an ideal tool to tackle combinatorial optimisation problems.

Quantum annealing is based on a crucial property of quantum systems: when in the ground state, the system remains in the ground state if its evolution is adiabatic (slow). A quantum annealer begins initialised to known state. The energy landscape (the Hamiltonian) of the system is then slowly evolved over time until it is made to match a function that we want to minimise. According to the laws of quantum mechanics, the system remains in its ground state, thereby evolving to the minimum of our function: the solution to our problem! If we can map our problem onto the energy landscape of a quantum system, then a quantum annealer can solve the problem for us.

This technology has recently become commercially available, with a handful of companies offering relatively accessible quantum annealing services. There is growing interest in their potential across many industries, from finance to pharmaceutical research.

At TTP, we’ve been keenly exploring possible applications for quantum annealing. One area in which we think it could be helpful is spatial biology.

Spatial biology is a relatively young field filled with excitement and promise – it was named Method of the Year 2020 by Nature Methods. For the first time, the technology now exists for scientists to study gene expression at the resolution of individual cells within a tissue. This is a huge step forward from previous approaches, in which tissues were often homogenised to collect averaged data.

Nevertheless, reconstructing tissue gene expression at cellular resolution is sensationally complex. A powerful approach consists in covering a target tissue slice with numerous beads, each with a unique DNA barcode and functionalised to collect nucleic acid fragments, i.e., the local gene expression profile. The beads are then removed from the tissue and the collected nucleic acids sequenced. The unique barcode still makes it possible to assign each gene expression profile to an original bead.

But how can you tell where each bead was in the tissue? Either you map each bead beforehand , which is expensive, or you can try to reconstruct the tissue layout using the fact that adjacent cells tend to have similar gene expression profiles.

TTP has developed a patent-pending approach to guide the reconstruction by mapping the position of a small subset of beads . This set of special beads – the “landmarks” – gives us a “low resolution” picture of our tissue. How low can we take the resolution of this, while still being able to reconstruct the “high resolution” picture accurately? Locating beads is costly, therefore, the smaller the number of landmarks, the better.

Classically, finding the best reconstruction – even given the low resolution picture – is computationally prohibitively difficult. Approximate algorithms are indispensable. But what if we don’t know how many landmarks we need for the low-resolution picture, or whether an approximate algorithm works? There are so many combinations that even “knowing what needs fixing” can be an immense challenge.

At TTP, we are applying quantum annealing to check what is the “lowest resolution” of the landmarks that still allows us to accurately reconstruct the high resolution gene expression picture across the tissue. Once we are reassured that the “low resolution” picture is good enough, that will give us confidence to fine-tune classical algorithms on larger datasets, in the knowledge that we are moving in the right direction.

In short, quantum annealing is well placed for developing and benchmarking classical algorithms in highly complex fields such as spatial biology.

Our Quantum Technology Team has been busily examining spatial biology among other potential fields in which quantum computing could be a game-changer. TTP is in a unique position to explore practical applications for quantum computers – we are proud to have a culture in which experts in quantum computing work side-by-side with experts in life sciences and countless other fields, exploiting a range of tools, and helping clients at every stage on the road to commercialisation. Our quantum computing expertise thrives in this multi-disciplinary environment.

We’re excited to continue searching for applications for quantum annealing and other quantum computing technologies. We can help examining whether they could be suitable for your challenges, and, if so, help you implement them.

To find out more, get in touch.