Terra Quantum has announced a breakthrough in molecular modeling through its collaboration with Professor Dr. Christoph Bannwarth of RWTH Aachen University. The team developed a quantum tensor network-based method that accelerates the molecular conformer search process by 5X to 20X compared to classical techniques, offering significant improvements in speed, accuracy, and flexibility. This innovation promises to streamline drug discovery and materials science applications, potentially reducing development costs by improving molecular structure prediction efficiency.
The study, published on ChemRxiv, demonstrates the method’s superiority over existing tools like CREST and its ability to explore larger chemical spaces without requiring extensive datasets. Unlike conventional AI approaches, Terra Quantum’s method does not need training data, making it particularly effective in limited-data environments. The new technique was tested on molecules like penicillin and ritonavir, achieving faster and more reliable results.
Professor Dr. Christoph Bannwarth emphasized the importance of the new tool for large-molecule analysis, stating, “This opens opportunities to treat large molecules that would otherwise lead to an explosion in computational costs.” The next phase of the research will focus on even larger molecules, including cyclic peptides, with applications in protein binding and drug development.
A press release announcing this development can be viewed here and the paper, titled, Tensor Train Optimization for Conformational Sampling of Organic Molecules, can be accessed here.
October 3. 2024
