Quantum Computing’s Role in Accelerating Drug Discovery
Industry leaders explore how hybrid AI–quantum approaches could unlock new medicines faster despite hardware and skills challenges
Quantum computing’s potential to transform drug discovery is gaining momentum, with leading voices from the pharmaceutical sector highlighting its ability to improve molecular modelling, optimize clinical trials, and identify new biomarkers.
Far from replacing AI, experts view quantum computing as a powerful partner in hybrid workflows that can deliver more precise, efficient, and innovative treatments.
For Agnes Meyder, principal scientist at Roche, the promise is significant but must be backed by tangible results within the tight timeframes that govern pharmaceutical R&D investment.
Working within Roche’s biochemical research pipelines, Meyder and her teams explore early-phase drug discovery—typically the first three to five years of identifying and shaping drug candidates.
She said this stage is full of “challenges that are classically unsolved in quantum chemistry and optimisation.” However, the internal reality is competitive: any proposed approach must outperform other high-performing technologies in terms of resources.
“If you make an inbuilt decision — putting up a team and bringing it in — you need to see success within the next three years,” Meyder said. “In our organisation, we are already seeing large language models (LLMs) win those internal pitches because they have demonstrated scalable results. Quantum has to show it can do the same.”
On the technical side, she emphasized that success hinges on encoding biochemical problems into quantum-compatible algorithms.
“Once you know how to translate your problem into the quantum framework, the learning curve becomes much less steep,” she said.
She pointed to the rapid adaptation of engineers to AI tooling as proof that “people of all ages and backgrounds can pivot quickly once the methodology is clear.”
Meyder also praised the role of shared resources.
“The base tooling might be known to enough players, small and big,” she said, “but how to use the tooling so that it fits your problem — that is the core of the trick.”
Blending Physics and Pharmaceuticals
The perspectives from Meyder formed part of a broader discussion at Commercializing Quantum Global 2025, held in London and organized by The Economist. The event explored how quantum computing could be integrated into pharmaceutical research and development (R&D), from early discovery through to trial design.
When the conversation turned to practical applications, Bill Prucka, associate vice-president - research - advanced analytics and data sciences at Eli Lilly and Company, said that “getting the calculation done quicker is going to be less valuable than making better molecules.”
He explained that much of Eli Lilly’s drug discovery work involves “calculating molecular properties and screening potential compounds.”
While AI has transformed these processes, most famously through breakthroughs like AlphaFold’s protein-structure predictions, Prucka said that “there are points where you need more exactness than what AI approximations can deliver. That’s where quantum will make a difference.”
Prucka argued that quantum could serve as an equalizer for smaller firms, enabling them to develop drug candidates quickly and potentially challenge the advantages of Big Pharma.
“If we’re not ready, someone else will be,” he warned.
He said that “the real challenge is not only in building the quantum hardware, but in bridging the gap between defining a research problem and solving it with quantum tools. If you can’t connect those dots clearly, you risk losing executive interest before the technology has a chance to deliver.”
Timelines and Managing Expectations
For Lene Oddershede, senior vice-president of natural and technical sciences at the Novo Nordisk Foundation, the timeline for quantum’s full impact in pharmaceutical R&D is likely six to seven years — if hardware advances as expected.
“Right now, our machines can maybe do 10⁴ error-free operations,” she said. “To run large-scale drug discovery, we’re going to need on the order of 10¹². That’s a huge leap.”
She said that “being realistic is really number one if you want to convince senior management.” When she joined the Foundation, she told its board that quantum impact was at least 10 years away.
“They said, ‘If it’s 20 years, that’s still fast,’” she recalled.
Oddershede highlighted molecular interaction modelling as one of quantum’s most valuable future contributions to pharma.
“We will be able to simulate the behaviour of molecules in a physiological environment with accuracy that classical computers just can’t match,” she said.
She also sees quantum sensing, which is already being used in hospital prototypes, as a near-term application for diagnosing metabolic diseases from biological samples.
She argued that first movers in quantum adoption may bear the cost of pursuing the wrong technical path. She added that “second movers can avoid those mistakes and jump in with proven methods.”
On hardware, she was unequivocal: “When we build the first industrial-scale, general-purpose quantum computer, that’s when the breakthrough happens. Until then, we are talking incremental improvements — better qubits, improved error correction, and hybrid algorithms that squeeze more out of what we have now.”
Small Firms, Bold Moves
While the Foundation’s long-term view allows for patient investment, the competitive landscape still demands agility. Oddershede acknowledged that smaller companies can sometimes move faster, demonstrating early benefits of quantum that larger organizations might then adopt or acquire.
Lara Jehi, chief research information officer at the Cleveland Clinic, a nonprofit health system, said: “We want to be ready for the technology whenever it’s ready to answer our biomedical research problems.”
She outlined a hybrid approach to biomarker discovery.
“We start with classical machine learning, move to AI-driven foundation models, and then layer in quantum machine learning,” she said. “Pitting quantum against AI is really a very artificial way of looking at it. By nature of where we are with the technology, you have to go back and forth between AI and quantum.”
She argued that quantum computing is a “Box Three” technology in the three-box solution framework — critical for future competitiveness, even if it does not yet generate immediate returns.
“Box One is the present, where AI dominates. Box Two is what you have to let go of to innovate. Box Three is the future — and if you’re not investing in Box Three, the world will pass you by,” she said.
Hardware: The Bottleneck and the Race
Despite different organizational contexts, all four speakers agreed on one point: quantum hardware is the main limiting factor. More reliable qubits and industrial-scale systems are essential for pharmaceutical applications beyond experiments and pilot projects.
Meyder said that “community collaboration will be essential for keeping momentum during this hardware bottleneck, particularly in refining problem-encoding strategies that can be shared across the industry.”
Prucka emphasized that “quantum will not have a single role in pharma; in some areas it will be a supporting tool, in others — such as quantum sensing — it may lead the approach.”
Jehi and Oddershede both said that hybrid AI–quantum workflows are the most realistic way forward. These approaches allow incremental improvements while preparing teams for the eventual arrival of large-scale, error-corrected quantum computers.
If that trajectory holds, the next generation of medicines could emerge from a research environment where AI’s predictive capabilities are fused with quantum’s computational precision.
As Meyder said: “When that happens, it won’t just be about doing the same things faster — it will be about doing things we simply couldn’t do before.”