Intelligent pharmaceutical patent search on a near-term gate-based quantum computer

Area of Science:

• Chemistry
• Quantum Computing
• Intellectual Property Law

Background:

• Pharmaceutical companies rely on patent analysis for product protection.
• Markush structures in patents represent chemical compounds with variable substituents.
• Identifying overlaps between diverse Markush structures is challenging due to their complexity.

Purpose of the Study:

• To develop a quantum search-based method for exact comparison of non-enumerated Markush structures.
• To demonstrate the feasibility and potential quantum advantage of this method in pharmaceutical patent analysis.

Main Methods:

• A quantum search algorithm utilizing a constraint satisfaction oracle was designed for Markush structure comparison.
• The quantum circuit was validated using a quantum simulator.
• Real-world noise effects were assessed on a five-qubit superconducting quantum computer (IBM).

Main Results:

• The quantum method enables exact comparison of complex, non-enumerated Markush structures.
• Quantum circuit performance was evaluated, identifying depolarizing error as the most significant.
• Improving qubit connectivity can enhance measurement accuracy.
• The method is computationally intensive for classical simulation, indicating a quantum advantage.

Conclusions:

• Quantum computing provides a powerful tool for precise pharmaceutical patent analysis.
• The proposed quantum method offers a significant advantage over classical approaches for complex Markush structure comparisons.
• Further research into quantum hardware and error mitigation can enhance the practical application of this technique.