A team led by Kenneth Merz, PhD, of Cleveland Clinic and Antonio Mezzacapo, PhD, of IBM has developed a hybrid quantum-classical computing model to simulate and study supramolecular processes that guide how entire molecules interact with each other. The research, published in Nature Communications Physics, focused on molecules’ noncovalent interactions, particularly hydrogen bonding and hydrophobic species, which are important in processes like protein folding and cell signaling.
The hybrid methodology leverages quantum-centric supercomputing to overcome the limitations of conventional quantum methods, which often lack the required accuracy for simulating complex noncovalent interactions. Researchers used an IBM Quantum System One to generate samples of different possible molecular behaviors for two supramolecular systems: water dimer and methane dimer. The classical computer then processed these samples to output chemically accurate molecular energies.
Dr. Merz noted that the hybrid models can significantly reduce the time and cost of computation while solving scientific bottlenecks. This approach, which uses the Sample-based Quantum Diagonalization (SQD) method, is intended to accelerate the discovery of new treatments and drugs. This work builds on the accelerator’s prior efforts to apply hybrid models to chemistry, including a project that explored using quantum machine learning for predicting proton affinities of molecules.
Read the full announcement from the Cleveland Clinic here, the study in Nature Communications Physics here, the IBM technical overview here, and the related QCR article on predicting proton affinities here.
November 23, 2025
Leave A Comment