With Electronic Structure Descriptors and Artificial intelligence
Today, most chemicals are still assessed using methods developed decades ago, or tested in animals. Computational toxicology was meant to change that, but existing models hit a wall: they only work well for molecules similar to those used to train them.
Quantum-Tox wants to tear down that wall. By anchoring predictions in quantum mechanics – the physics that governs how electrons actually behave inside a molecule – our models can, in principle, assess any molecule, even ones that have never been synthesised.
a future where computational toxicology is centred around electronic signatures – compact, quantum-mechanics-based descriptors that capture the true behaviour of a molecule from the inside out. Interpretable by chemists, scalable by AI, and reliable across the entire chemical space.