How Quantum Could Deliver Faster, Safer, Better Drugs: Q2B 2023 Paris

The life sciences market is often cited as likely to be one of the first to achieve quantum advantage, but results from current noisy, intermittent-scale quantum (NISQ) computers have at best matched classical computing. However, progress in the field is proving promising.

A panel at Q2B Paris assembled quantum experts from the pharmaceutical industry to discuss the current state of the art and how their organizations are looking to adopt quantum into their processes.

Quantum Use Cases for Pharma

Xavier Cimino, a partner at management consultancy McKinsey specializing in AI and quantum technologies, chaired the panel and started by setting the scene. He said pharma is one of the strongest players when it comes to innovation, and in the last 10 years, it has spent more than $1 trillion on research and development. That investment has translated into the launch of more than 500 completely new and innovative drugs.

“Innovation is at the core of what pharma does, and this will continue to help save patients’ lives,” Cimino said. “During the pandemic, R&D accelerated and we were able to launch vaccines in less than a year. The previous record was four years.”

He added that quantum has the potential to further accelerate R&D, for example by helping drug discovery and improving understanding of drug structure and the associated biochemical reactions to discover faster, better, safer drugs. But some challenges mean making it a reality will be quite complex.

Gopal Karemore, director of digital research intelligence at Danish multinational pharma company Novo Nordisk, said applying quantum computing starts with the same question with which pharma companies start any new research: which disease do you want to focus on?

“To understand that you have to have a mechanistic understanding of the disease progression and understand the replication, transcription and translation mechanisms around it,” he said. “We have a ‘multiomics’ approach that needs to take into account genomics, metabolomics and proteomics. This is not a question of big data; it is a question of the complexity of the problem. Quantum has some solutions for this.”

He added that the second part of the equation is the simulation of all configurations of a new candidate molecule, which quantum has proven theoretically to optimize. Those are then entered into a compound library, a collection of chemicals that can be used for high-throughput screening and other processes for drug development then the interaction with combinations of other molecules. Quantum could help with both processes.

When a drug is approved, quantum can optimize the clinical trials process and production and the supply chains that bring it to market.

“Last but not least, is quantum machine learning,” said Karemore. “Classical machine learning is already integrated ubiquitously into the entire drug discovery and development pipeline, so quantum machine learning holds some hope. Right now, we use a pseudo-random number with a classical computer, so the hope is that we can use a quantum random number generator.”

How to Engage Decision Makers with Quantum

While these potential use cases are readily accepted among the quantum community, getting them accepted by scientists and executives may prove a harder sell. Cimino asked Jens Kieckbusch, associate director of emerging technologies at AstraZeneca, how to create a level of adoption of excitement with both scientists and executives.

“I work in a part of the business that's called the emerging innovations unit. And we bring in the outside world when it comes to innovation, looking for technologies, projects and new ways of doing things that have an impact horizon on patients on a five to 10-year timescale,” Kieckbusch said.

“That usually means new indications and new ways of measuring cells, but it also means cross-cutting technologies such as quantum. There is always a spectrum of people, with the early adopters on one hand and the healthy skeptics on the other. I think it's really important that you have a team internally that knows who these early adopters are because you need a little bit of their time to develop these proof-of-concept studies for emerging technologies to generate the positive data that will then hopefully convince the skeptics that this is something that we should do.”

Kieckbusch added that pharma scientists are usually busy, working on time-dependent tasks that could directly impact patients, so it’s a difficult value proposition. While focusing on early adopters usually works well, it requires people who know their way around the R&D organization and are well-connected within the company. There is also a need for educating decision-makers as the public domain news about quantum tends to be dominated by hype. When it comes to the decision

“We need a team that can change the narrative internally and says, yeah, it’s a bit too early, but out of these 39 use cases we found this one where we could do a little bit of work, maybe work with an external partner, and start a small program of work to do proof of concept to see how this can deliver in the future,” said Kieckbusch.

“You need to cut through this negativity that you get often with emerging technologies and make a clear business case setting out an opportunity rather than just forwarding papers or conveying information. Be honest, generate the data and convince people with evidence.”

Quantum Challenges

Clemens Utschig-Utschig, head of IT technology strategy at pharmaceutical company Boehringer Ingelheim went on to point out that developing sufficiently powerful hardware isn’t the main problem quantum faces,

“Everybody seems to believe quantum is a hardware problem,” he said. “Suppose an alien starship flies by and drops a quantum box with 10,000 logical qubits on us. We would treat it as if we were in the stone age and someone discovered fire and said, whoa, that's warm! It will take years to put algorithms on this beast. It’s not like you take an algorithm or a problem that you know today and say awesome box there, let’s go spin up some qubits.”

Utschig-Utschig gave the perspective of a recent academic paper that took a year and a half, involved talking to tens of people and totaling thousands of hours of work, saying that is the scale of work that faces quantum research.

“On the algorithmic side, there's a lot to do, and then there's all the other things around it, like how would you embed all that? How do we get information out of the box? What type of information will you get out? Wave function, data point, anything in the middle? You need to create some sort of state that you can put on the kind of box right and that state is rather pretty close hopefully to what you want to get out and someone needs to create that state. How do we efficiently create that state? It’s not that I'm a pessimist, these are challenges and hard research questions.”