Qubit Pharmaceuticals, a deeptech company specializing in molecular simulation and modeling through the use of quantum physics, announced on Wednesday, November 30, the creation of a platform to accelerate drug discovery using hybrid classical and quantum computing on the NVIDIA Quantum Optimized Device Architecture (QODA).

Qubits Pharmaceuticals was founded in 2020 and its five co-founders are from leading research institutions in France and the United States: CNAM, CNRS, University of Texas at Austin, Sorbonne University and Washington University in St. Louis.

The company, based in Paris at the Paris Santé Cochin incubator and in the U.S. in Boston, is a member of NVIDIA Inception, a program that provides business support, expertise, and go-to-market technology for cutting-edge startups. It has a 25 pflops supercomputer, Gaia, one of the largest GPU supercomputers in France for drug discovery.

It has also developed a software suite, Atlas, which, by creating true digital twins of physical molecules, performs in a few hours calculations that would take several years by conventional means, i.e. an acceleration of a factor of 100,000. While a classical research requires an average of 5,000 molecular syntheses, the simulation reduces this number to about 200 and also allows to explore several chemical families in parallel, rather than sequentially.

Thus, it is no longer necessary to synthesize drug candidates developed in chemistry or using artificial intelligence algorithms to validate their efficacy: Atlas allows the modeling of molecules, their description and therefore the prediction of their action.

Atlas has enabled the company to discover new drug candidates against Covid in less than 6 months and to initiate several research programs in the field of cancer and inflammation.

Improving modeling and simulation of complex targets

Drug discovery through simulation and modeling requires unprecedented computational capacities exceeding one petaflop, or more than one million billion operations per second.

For Qubits Pharmaceuticals, the future of molecular modeling and simulation in drug discovery should lie in the hybrid use of classical supercomputers and quantum computers, with quantum computers serving as gas pedals for certain key calculations alongside GPU supercomputers.

To explore this new area of research, Qubit has chosen NVIDIA QODA, a hybrid quantum-classical computing platform. It enables the seamless integration of quantum computing with GPU-accelerated supercomputing, and is compatible with all quantum resources, including physical QPUs and emulation with cuQuantum.

QPUs, or quantum processing units, could accelerate key software kernels for molecular modeling. The NVIDIA cuQuantum SDK, meanwhile, allows Qubit developers to simulate quantum circuits and thus design algorithms ready to run on future quantum computers.

Qubits’ new platform is accelerated by 200 GPUs (next-generation graphics cards) including NVIDIA DGX A100 systems, offering unparalleled power for AI simulation and computation, according to the company. The combination of NVIDIA’s accelerated computing platform and Qubit’s Atlas software aims to cut the time required to screen, select and optimize a drug candidate of interest by more than a factor of 2 and the investment required by more than a factor of 10.

Initially, this new platform aims to develop 10 internal research programs in the fields of oncology, inflammatory diseases and antivirals. The first drug candidates should be tested by pharmaceutical companies as early as 2023.

Robert Marino, CEO of Qubit Pharmaceuticals states:

“We are extremely proud two years after Qubit’s inception to have earned the trust of a global leader like NVIDIA. By combining NVIDIA’s computational power with Qubit’s simulation and molecular modeling capabilities, we are confident in our ability to dramatically reduce drug development time and cut the cost of pre-clinical development by a factor of ten. This unique collaboration is also groundbreaking in the field of quantum physics and should enable us to develop the first algorithms in this field applied to drug discovery.”

Translated from Qubit Pharmaceuticals s’appuie sur NVIDIA et l’informatique quantique hybride pour accélérer la découverte de médicaments