Qubit Pharmaceuticals Uses Quantum AI To Solve Complex Chemistry Problems

This new technology aims to speed up drug discovery while cutting costs significantly.

Developed in partnership with Sorbonne University, the model can simulate and predict how molecules behave with unmatched accuracy and speed.

One of the toughest parts of drug discovery is predicting how well a drug will bind to a protein, RNA, or DNA. Since there are trillions of possible molecule-target combinations, it’s impossible to store them all in a database.

Read More - Sensmore Raises $7.3M Fund For Fuel Autonomous Industrial Vehicles

Qubit’s quantum AI helps solve this by accurately predicting these interactions, reducing the need for expensive lab testing and chemical synthesis of drug candidates.

“This simulation method would reduce the cost of the drug discovery phase enormously,” says Jean-Philip Piquemal, Professor at Sorbonne University and Director of the Theoretical Chemistry Laboratory (Sorbonne University/CNRS), co-founder and Scientific Director of Qubit Pharmaceuticals.

“The model is as accurate as the experiment; we can generate many new ideas, failing quickly and cheaply ‘in silico’ before moving on to laboratory testing with molecules that have passed the tests with flying colours.”

The team used powerful computing resources from GENCI, EuroHPC, and Argonne to build FeNNix Bio1 — a foundation model based on millions of detailed molecular simulations.

They trained the model using what they call the world’s most accurate chemistry database, created with high-precision simulations. This helped the model learn the basic rules of chemistry and physics. Like building with Lego blocks, the model can now piece together complex biomolecules and understand how they interact with each other.

FeNNix-Bio1 has shown strong results in one of the toughest challenges in molecular modeling: simulating how water behaves in different forms.

The model can accurately predict physical properties and replicate how ions and small organic molecules behave in water — something many other models struggle with.

This matters because water is the main solvent in the human body, and how it interacts with drugs is crucial to how those drugs work.

The Sorbonne University team behind FeNNix-Bio1 aimed to push beyond what Google DeepMind’s AlphaFold can do. While AlphaFold predicts protein structures based on amino acid sequences, FeNNix-Bio1 takes it a step further, offering deeper insights and broader capabilities.

According to Jean-Philip Piquemal, AlphaFold has revolutionized protein structure prediction. “However, proteins are not static; their structures evolve over time, modifying drug interactions. FeNNix-Bio1 makes it possible to model these dynamic effects.

In addition, AlphaFold does not accurately model the interactions of proteins with drug candidates. FeNNix-Bio1 addresses these two important limitations for biomolecular simulation.”

Traditional simulations often lack precision or speed, and while quantum chemical models are accurate, they’re too slow and expensive for large-scale use. FeNNix-Bio1 changes that—it delivers quantum-level accuracy while staying scalable and affordable. Unlike models that just predict molecular structure, FeNNix-Bio1 understands how molecules behave and interact.

Instead of relying on language model (LLM) architectures built for text, the team designed custom neural networks specifically for chemistry and physics.

FeNNix-Bio1 is also faster and cheaper to train. It can be trained in just a few hours on a regular GPU, while other models might take weeks on powerful supercomputers.

“We're aiming at complex targets, those for which the pharmaceutical industry doesn't provide a solution for patients,” comments Robert Marino, CEO of Qubit Pharmaceuticals.

The company is currently working on 7 drug discovery programs, mainly focused on cancer and inflammation. The most advanced one targets breast cancer.

Since the model is based on the fundamental laws of physics, it’s highly flexible—by adjusting the molecular building blocks, it can simulate virtually any system.

Beyond drug discovery, the technology can also be used in the chemical industry for designing industrial enzymes, improving desalination membranes, developing next-gen batteries, and speeding up green chemistry innovations.

FeNNix-Bio1 is also a step toward quantum AI — combining quantum computing with machine learning to transform how molecular data is generated.

According to Piquemal, Qubit Pharmaceuticals is already using quantum data to enhance its models — a breakthrough many believed wouldn’t be possible until 2035.

About Qubit Pharmaceuticals

Qubit Pharmaceuticals is a drug discovery company that uses physics-based methods to develop new treatments. It focuses on difficult targets and designs drug candidates atom by atom to boost effectiveness, precision, and safety. Founded in 2020, the company is based in Paris, France.