At a glance . . .
We use Artificial Intelligence to navigate through the immense chemical universe, to identify islands of interesting and novel chemical matter, and to use complex computer simulations to optimize both, the chemical reactions leading to promising compounds and all their properties that make them the drugs of the future.
Our platform combines AI technologies for disease and target selection, drug design, novelty assessment, complex decision-making, and evolutionary learning.
Where AlphaGo has managed to learn the few given simple rules of Go to beat the human world champions, our AI platform learns rules from all available chemical knowledge and applies it to design the most suited small molecules for a given protein target or – likewise – to identify the most suited protein target for a given small molecule.
We link AI based drug design to efficient wet-lab realization, property measurement, and biological characterization, thus creating a self-learning feed-back cycle.
Origenis explores new chemical spaces where no human has gone before. Our scalable AI platform designs novel chemical matter and optimizes them in a repeatedly proven process towards synergistic pipelines of high-quality preclinical development candidates.
Our worldwide patent knowledge and analytics technology learns constantly with weekly updated patent applications and granted patents ranging back to 1970. Cippix® analyzes in real-time compound, target and disease information from industry relevant patents, uncovers current research strategies, future development focus and planned application and use. Cippix® turns unstructured patent data into the largest and most up-to-date prediction engine.
Our AI-driven drug design and synthesis planning engine combines soft-synthesis planning fed by >150,000 reaction rules applied to existing in-house and commercial building blocks, state-of-the-art prediction for activity, selectivity and drug-like properties needed for the planned therapeutic indication. These rules are automatically learned with Onion™-technology and encompass all kinds of organic and biochemical reactions. Parallel evolutionary strategies applied to ultra-large chemical spaces and drug design enable the selection of best suited and synthesizable compounds from more than trillions of trillions compounds not yet described or enumerated.
The unique and seamless transition from AI driven design to realization. MolMind® planned compounds are synthesized and screened to generate real-life data, thus, combining the in-silico design with real world data. Feeding back novel and proprietary data to Origenis’ AI engine is resulting in enhanced synthesizability and optimized activity and selectivity.
Using our proprietary technology, we assess important physicochemical properties of all of compounds synthesized using only fractions of a milligram per compound. This enables us to calculate tissue distribution e.g. blood-brain distribution or eye-availability already very early start in a program ensuring the best-possible optimization of the compounds for their planned application. Results of BRAINstorm™ are constantly fed back to our in-silico models to understand the structure-property-relationships not only for the whole molecules, but atom by atom to sharpen and expand the property prediction within MolMind™.