Using quantum computing in life sciences

In a recent article, McKinsey Digital estimated the value of the opportunity to big pharma of applying quantum computing to protein structure and interactions at $200 billion, and highlighted that the opportunity across a range of technology-rich industries far exceeds that.

This potential is reflected in recent partnering announcements in the life sciences space, including:

• IBM and the Cleveland Clinic entering into a 10 year partnership to accelerate discovery in healthcare and the life sciences
• Boehringer Ingelheim and Google collaborating to apply quantum computing to molecular dynamics simulations in a dedicated Quantum Lab
• Caltech partnering with Amazon to create a new Quantum Computing Hub
• Capgemini becoming an IBM Quantum hub.

So why life sciences and why now?

As far back as 1982, Richard Feynman presciently wrote: "if you want to make a simulation of nature, you'd better make it quantum mechanical". Feynman recognised that the underlying processes which govern nature are quantum mechanical, so it follows that in order to simulate nature reliably in silico, the model will also need to be quantum mechanical.

While the underlying theories are well-understood, even with four decades of exponential growth in computing power, the state of the art in computation still struggles to simulate the multiple complex biological interactions that come together in order to define, for example, the activation of a target by a drug molecule.

It is in these sorts of use cases that quantum computing has the greatest potential to disrupt. It is not simply a matter of those use cases being the applications which push the limits of current computational power. The nature of the underlying processes are such that algorithms that rely on superposition – the ability of qubits to be both a 1 and a 0 at the same time – are needed in order to produce an efficient and accurate model, where existing computational techniques struggle for power.

The life sciences arena – pharmaceutical R&D in particular – is replete with fields of research that are potential use cases for quantum computing. These include molecular design, molecular similarity, protein folding and protein-ligand interactions, modelling mechanisms of drug action, biomarker discovery, quantitative structure activity relationships, and modelling the behaviour of larger biological systems. More broadly, pathology and image analysis, the fields of precision and personalised medicines and of genetics (particularly linking genomes and outcomes), all involve complex systems where the deployment of quantum computing could have a game-changing effect.

Looking at the pharma space, while computational techniques are ubiquitous, there is still a pressing need to reduce time and cost to market, increase efficiency and reduce attrition rates. It is difficult to envisage this need being met by advances in transistor-based computing power alone. There is a compelling case that quantum computing will lead to a paradigm shift, with recognition of that potential being borne out by the year-on-year increases in the value of private investment into quantum computing as we discuss here.

Staying flexible

An interesting facet of the development of quantum computing hardware is the variety of different technological approaches to generating qubits. While it is unclear which technology or technologies will 'win' (ie which will give rise to a scalable quantum computer with acceptable levels of errors and capable of solving complex real-world problems) the multiplicity of hardware has prompted a need for 'platform agnostic' software, fostering an ecosystem of developers of platforms and algorithms that are capable of running on any quantum computing hardware.

This disconnect between hardware and software could lessen the time taken for the potential of quantum computing to be realised. Instead of advances in quantum computing power leading to new applications, the applications are being developed in parallel to the quantum computers themselves. Ground-breaking algorithms could be ready to run as soon as the requisite quantum computing power becomes available.

Getting quantum ready

Deploying quantum computing in the life sciences will involve the coming together of hardware developers that create the first successful scalable quantum computers, developers of algorithms and software platforms to enable quantum computers to be deployed towards solving real-world problems, and those engaged in R&D, such as pharmaceutical companies and those in other technology-rich areas, who are capable of translating the output from the use of quantum computing techniques into real-world products.

Each of those parties has a different skill set and will need to adjust their thinking in order to capitalise on the opportunities offered by quantum computing. The trailblazers will not be those who simply substitute a quantum computer to do a job that is currently done using existing computing capabilities (albeit with a step change in computing power). Collaborations in the life sciences with require an understanding both of what quantum computers can achieve and of the underlying life sciences processes being modelled in order to capitalise on the potential of quantum computing techniques.

Several years ago, advances in the development of AI systems and of computing power made the practical deployment of AI towards solving complex problems in the life sciences a realistic possibility. This prompted substantial investment and significant partnering activity, including between major pharmaceutical companies and the leading players in the AI space. Quantum computing seems to be following a similar trajectory, with systems capable of practical deployment in R&D being a realistic possibility in the near term.

Those looking to capitalise on practically useful quantum computing solutions as and when they become available will already be making preparations – referred to as becoming 'quantum ready'. The expectation is not that quantum computing will immediately replace existing methods, but rather that there will be a process of bringing quantum technologies onstream as and when their deployment becomes beneficial.

Becoming quantum ready therefore involves plotting a path from the use of existing technologies and being able to pivot to quantum technologies once available. Given that platform agnostic quantum software is already a reality, prospective early adopters have visibility of what the quantum future looks like and can plan their quantum computing strategy and make resourcing decisions accordingly.  This has already been borne out by investor and partnering activity - a trend that looks set to continue. It is only a matter of time before significant success stories start making the headlines.