Quantum software matches classical benchmark for a molecular interaction
Predicting how molecules interact is central to drug discovery: it offers important information about how a new drug might work in the human body, the side effects it can cause and more. Computationally modeling these interactions can reduce the need for lengthy and expensive lab work, but some interactions still remain beyond the reach of classical computers.
Predicting how molecules interact is crucial in drug discovery, as it provides insights into a drug's functionality, potential side effects, and more. Classical computers struggle with some molecule interactions, but Danish quantum software company Kvantify has made progress in overcoming this limitation. By employing a novel computational model tested on Quantinuum's System Model H2 quantum computer, they successfully captured a dispersion interaction relevant to biomolecular stability.
Dispersion interactions, though small individually, significantly contribute to stabilizing biological structures like DNA and proteins and influencing molecular binding. Modeling these interactions is challenging due to the complex behavior of electrons, which requires extensive computational resources. Kvantify's quantum algorithm, BEAST-VQE, combined with exact quantum circuit simulation using Qrunch's optimized integration with NVIDIA cuTensorNet, demonstrated the accuracy of computational predictions rather than claiming commercial quantum advantage.
The study used a simplified model, covering two geometries instead of the full interaction curve, but its significance lies in showcasing the value of integrating quantum and classical methods for molecular interaction investigation, even with current quantum computers. The work also underscores the need to address classical computational bottlenecks as higher-quality quantum hardware becomes available.
BEAST-VQE matched the CCSD(T) accuracy for dispersion interactions in a simplified DNA-derived molecular pair, reducing active orbitals from 236 to 53, with quantum hardware results closely matching ideal simulations. This benchmark demonstrates accuracy rather than commercial quantum advantage.
Written by urgent.news from Phys.org's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.