The rapid advancement of AI-powered tools, such as RoseTTAFold and AlphaFold, has revolutionised protein structure prediction, creating exciting opportunities across drug discovery and structural biology. However, while deep machine learning (ML) has delivered significant gains in static structure prediction, it remains limited when simulating the dynamic, atomic-scale interactions that determine true protein function. This is where quantum computing comes in: a radically different paradigm that, though still under development, could eventually enable the high-fidelity modelling of molecular behaviour at an atomic level.
While most use cases remain theoretical, early-stage collaborations between quantum computing companies and pharmaceutical companies are exploring how quantum platforms might enhance protein modelling, molecular docking and energy-state simulations. If successful, this could fundamentally alter how drugs are designed, not by replacing AI, but by complementing it with physics-derived accuracy out of reach of classical computers.
Go deeper with GlobalData
Quantum computing 101: why it is different
Unlike classical computers, which process information in binary bits (0 or 1), quantum computers use qubits, which can exist in superposition (both 0 and 1 simultaneously). This allows them to explore many possible solutions at once, dramatically increasing their computational power for certain complex problems. They also leverage two unique quantum phenomena: entanglement, which links qubits in a way that allows information to be shared between them, and quantum tunnelling, which enables systems to pass through energy barriers that classical computers cannot. Together, these features allow quantum computers to explore vast numbers of potential solutions simultaneously, making them well-suited to complex optimisation and simulation tasks.
In drug discovery, this is particularly relevant in:
- Quantum chemistry simulations, when calculating energy states and electronic configurations of molecules
- Modelling protein-ligand binding interactions with increased precision
- Predicting dynamic folding/unfolding pathways and structural rearrangements for conformational analysis
While today’s quantum hardware is still limited by qubit stability (decoherence) and high error rates, ongoing advances in quantum error correction, qubit scaling and hybrid quantum-classical algorithms are rapidly closing the gap between theory and application.
Why protein modelling matters and where classical tools fall short
The accurate modelling of proteins is foundational to drug discovery. From understanding disease development and progression to designing ligands with optimal binding affinity, structural biology underpins modern medicinal chemistry. While tools like AlphaFold3 and RoseTTAFold have excelled at predicting static protein conformations, they cannot simulate conformational flexibility, rare structural transitions or quantum-level interactions such as proton tunnelling and charge delocalisation. These are important features that often determine binding kinetics, allosteric regulation and resistance mutations, all crucial for precision drug design.
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataClassical computers struggle to handle the complexity of electron interactions in large biomolecules due to the exponential growth in variables as molecular size increases. Quantum computing, by contrast, is inherently suited to modelling quantum systems, offering the exciting possibility of simulating proteins and molecular interactions at a resolution that classical physics cannot achieve.
Early collaborations
Several pharmaceutical companies have begun exploring quantum computing through research partnerships and pilot projects. For example, Boehringer Ingelheim collaborated with Google Quantum AI in one of the first public pharma-quantum collaborations, focused on simulating quantum chemistry problems relevant to drug design. Roche and Quantinuum have collaborated on quantum natural language processing and quantum-enhanced modelling of molecular interactions, while ProteinQure is a Canadian biotech building a proprietary platform combining quantum computing and machine learning for structural modelling and protein engineering.
These partnerships remain exploratory, but they reflect growing interest in quantum’s potential to complement existing AI/ML pipelines, especially in areas like conformational state prediction and ligand docking.
Protein folding and beyond: where quantum might help
Although AI systems such as AlphaFold have made remarkable progress in predicting static protein structures from sequence, they rely on statistical learning from known structures and often struggle with predicting alternative conformations and folding pathways, modelling membrane proteins and intrinsically disordered regions, and estimating binding energetics and thermodynamics in small-molecule interactions.
Quantum computing may offer a path forward by enabling alternative techniques such as:
- Energy landscape mapping, which allows for the determination of the most stable conformations and transition states
- Quantum-enhanced docking, which models binding affinity with greater precision using quantum Monte Carlo simulations
- Allosteric modulation prediction, which helps understand how distant binding events affect active sites
Though these applications are not yet clinically validated, they align well with the long-term goal of improving drug selectivity, potency and resistance profiles through more accurate biophysical modelling.
Current limitations: a technology still in the lab
Despite the hype, quantum computing remains constrained by several challenges.
Firstly, high error rates require robust error correction, which consumes qubit resources by using many physical qubits to protect a single logical one. Current hardware options limit qubit numbers and coherence time for large-scale protein simulations, and simulating even small proteins at high resolution requires hundreds to potentially thousands of logical qubits, depending on the complexity of the system and simulation method – orders of magnitude beyond today’s capabilities. Furthermore, there is a shortage of scientists who understand both quantum computing and molecular biology, meaning the software is still developing.
As a result, most real-world applications today focus on quantum-inspired algorithms or hybrid systems, combining classical machine learning supplemented with quantum computation to accelerate specific parts of the modelling workflow.
Quantum as a long-term strategic bet
For now, quantum computing is not replacing AI or classical simulation, but it is emerging as a complementary technology, especially valuable in tackling computational bottlenecks in molecular design. Its most realistic role in the near future will be as an adjunct to classical and AI tools, improving accuracy in areas such as high-energy conformations, rare protein states, and binding affinity prediction.
Over the next five to ten years, as hardware improves and error rates fall, we may see quantum computing become a critical component of structure-based drug design, which is especially exciting for undruggable targets with complex conformational behaviour, antibiotic resistance mechanisms that depend on protein flexibility, and allosteric inhibitors where small changes have large functional effects.
From experiment to reality
Quantum computing in protein modelling is still at an early stage, but it is no longer theoretical. Pilot studies, cross-sector partnerships and growing investment suggest that pharma is taking the field seriously. While full-scale quantum protein simulations remain a distant goal, hybrid approaches are already being explored, and the infrastructure is maturing rapidly.
For forward-looking research and development teams, now is the time to invest in quantum literacy, explore proof-of-concept collaborations, and monitor quantum’s integration into structural biology toolkits. If the next leap in drug design comes from modelling nature more as nature behaves – that is, considering quantum mechanics – then the industry must be ready to follow where physics leads.
