Due to its exceptional computing capabilities, quantum computing can revolutionize several scientific domains. Drug discovery, which requires predictive analytics and meticulous molecular modeling, is one field in particular that can significantly benefit from quantum computing.
Recently, efforts have been made to integrate quantum computing into drug design studies, which marked progress in applying advanced computational technologies for drug discovery. Yet, the use of quantum computing has been primarily limited to conceptual validation/proof-of-concept studies, which do not capture the drug development complexities in the real world.
The Proposed Approach
In this work, researchers established a hybrid quantum computing pipeline to address real-world drug discovery problems. They tackled two key challenges in computer-aided drug design: molecular dynamics simulation and computing reaction barriers. Their objective was to accurately simulate covalent bond interactions and determine the Gibbs free energy profiles for prodrug activation involving covalent bond cleavage.
Preparing the molecular wavefunction on a quantum device is crucial for the quantum computation of molecular properties. Thus, the proposed pipeline combined quantum-classical hybrid computing platforms, leveraging the variational quantum eigensolver (VQE) framework due to its suitability for near-term quantum computers to efficiently manipulate and store molecular wavefunctions.
The VQE core employed parameterized quantum circuits to determine the target molecular system’s energy. A classical optimizer was then used for energy expectation minimization until convergence. Consequently, the quantum circuit’s state approximated the target molecule’s wavefunction, and the measured energy was the variational ground state energy due to the variational principle.
Researchers used the ddCOSMO solvation model and analytical CASCI force formula to compute the solvation energy and molecular forces for quantum mechanics/molecular mechanics (QM/MM) simulation. The interface between classical and quantum computing relied on one- and two-body reduced density matrices.
Case Studies:
Two case studies using a superconducting quantum device demonstrated the pipeline's potential.
-
First Case Study: The Gibbs free energy profile was studied under solvent conditions for prodrug activation involving carbon-carbon bond cleavage. This strategy explores a novel, experimentally validated prodrug activation approach applied to β-lapachone for cancer-specific targeting. The prodrug design aims to overcome active drugs’ limitations in pharmacodynamics and pharmacokinetics, providing a critical supplement to current prodrug strategies. Accurate solvation effect modeling, crucial for simulating the prodrug activation process, was achieved using a general pipeline that implements PCM-based quantum computing of solvation energy.
-
Second Case Study: A covalent inhibitor for Kirsten rat sarcoma viral oncogene (KRAS) (G12C) was investigated using QM/MM simulation. The evolution of energy was closely monitored, and the time cost was compared between quantum and classical computers. KRAS, prevalent in several cancer types, plays a critical role in the RAS/MAPK signaling pathway, influencing cell survival, growth, and differentiation. Mutations of the KRAS protein, specifically the G12C variant, are common in cancers such as pancreatic and lung cancers. Sotorasib, a covalent inhibitor targeting this protein mutation, has shown potential for more specific and prolonged interaction with the KRAS protein. Quantum computing can enhance the understanding of these target-drug interactions through QM/MM simulations, which are crucial in the post-drug design computational validation phase.
Study Significance
Based on the results of the two case studies, researchers demonstrated the potential of the proposed hybrid quantum computing pipeline to solve real-world drug design problems. They highlighted the plug-and-play advantages and versatility of the pipeline.
Additionally, the effectiveness of quantum computing in scenarios involving Gibbs free energy profile calculations for covalent bond cleavage in prodrug activation was displayed. The time taken to calculate the energy expectation was 40 seconds. Furthermore, the calculation of the one-body reduced density matrices in the active space involved measuring three additional expectation values. Thus, the overall quantum computing kernel time cost was approximately 60 seconds, which was consistent with experimental findings.
A strong and specific bond between the target mutation and Sotorasib was observed, providing important insights into the potential efficacy of the drug. This understanding is critical for the rational design of future inhibitors targeting similar mutations.
Journal Reference
Li, W.et al. (2024). A hybrid quantum computing pipeline for real world drug discovery. Scientific Reports, 14(1), 1-15. DOI: 10.1038/s41598-024-67897-8, https://www.nature.com/articles/s41598-024-67897-8
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.