Quantum Research Development: Terra Quantum and RWTH Aachen University introduced a new method for molecular modeling using tensor networks, improving efficiency and accuracy.
Speed Improvement: The approach demonstrated a 5X to 20X increase in speed for predicting molecular structures over traditional methods.
Applications in Drug Discovery: The new technique could accelerate the drug design process, cutting down the time needed to identify viable drug candidates.
Terra Quantum, working alongside RWTH Aachen University, has developed a new approach for molecular modeling. Their tensor network-based technique has shown a significant boost in speed for determining molecular structures, with improvements ranging from 5X to 20X over existing methods. This collaboration, led by Prof. Dr. Christoph Bannwarth, has opened new doors for practical applications in both the materials science and pharmaceutical sectors.
The research has been published in a paper titled “Tensor Train Optimization for Conformational Sampling of Organic Molecules” on ChemRxiv. This discovery allows for more efficient exploration of complex molecules without the need for training datasets. By bypassing traditional approaches, Terra Quantum’s method aims to improve the entire molecular analysis process, providing a faster, more accurate understanding of molecular behavior.
Markus Pflitsch, CEO and founder of Terra Quantum, emphasized the potential this technique brings to the drug discovery process. By exploring a broader chemical space, pharmaceutical companies can identify promising drug candidates faster, significantly reducing the time and cost associated with drug development. This method holds promise for accelerating the search for stable molecules that can be used in pharmaceuticals and new material design.
"Being able to explore a wider chemical space could allow for significant acceleration in the drug design process by reducing the time needed to find a drug candidate most likely to be successful." – Markus Pflitsch, CEO, Terra Quantum
Terra Quantum’s approach focuses on improving the "conformer search" method, a technique used to determine the stable 3D structure of molecules. Unlike machine learning methods that require large datasets, this method operates efficiently without training data. By reducing computational costs and time, it presents a scalable option for industries that rely on detailed molecular analysis.
Researchers tested this method on a wide range of molecules, including well-known compounds such as penicillin and ritonavir. The process addresses the challenge of exponentially increasing conformations in molecules with many flexible bonds. Terra Quantum’s approach succeeds in mapping these structures without requiring massive computational power, demonstrating a practical solution for real-world molecular studies.
This research brings significant value to industries that rely on molecular modeling, including pharmaceuticals and materials science. Being able to predict molecular structures with increased speed allows researchers to accelerate the development of new drugs and materials. Given the high cost of drug development, with each approved drug costing an estimated $1 billion to $2 billion, the ability to streamline these processes will likely bring substantial cost savings.
Unlike conventional methods, Terra Quantum’s approach does not require the synthetic production of molecules to study them. This offers a data-independent method for understanding molecular behavior, allowing for more efficient analysis without the need for complex laboratory procedures. Researchers are now looking to expand the scope of this technique to larger molecules and more complex chemical structures.
The next phase of Terra Quantum’s research will focus on larger molecules, including cyclic peptides, and will explore protein binding affinities. The company will also be working with industry partners to apply the technique to specific, real-world needs. By continuing to enhance their tensor network technology, Terra Quantum is setting the stage for broader adoption of this method in molecular modeling.
This new technique also brings a physics-driven approach to molecular modeling, providing a more efficient solution compared to machine learning-based methods. The method eliminates the need for extensive datasets, allowing researchers to conduct predictive quantum chemical investigations without escalating computational costs. This breakthrough marks a critical step in improving how molecular analysis is performed.
Terra Quantum Group is a leading quantum technology company based in Germany and Switzerland. It provides various quantum services, including "Quantum Algorithms as a Service" and "Quantum Computing as a Service," offering powerful quantum computing solutions to businesses across a range of industries. Terra Quantum’s expertise enables companies to address complex challenges in logistics, optimization, and security using quantum technology.