Polish computational chemistry company Molecule.one has raised $ 4.6 million to expand its quest to bring theoretical drug molecules to life. Its machine learning systems predict the best ways to synthesize potentially valuable molecules, a crucial part of the creation of new drugs and treatments.
Molecule.one took to the stage at Disrupt SF 2019’s Startup Battlefield, where they explained the challenges faced by the drug discovery industry, essentially that they offer many theoretical treatments, but cannot actually make the molecules.
The company’s system kicks in when you have a new exotic compound that you want to craft so that you can test it in real life, but you don’t know how to do it. After all, these molecules are all new to science – no one has created them before, so why should anyone know? Molecule.one creates a workflow from common chemicals and provides step-by-step instructions using known methods to go from A to B… and to C, D, etc. (it’s rarely easy).
The company relies on machine learning and a vast body of knowledge about chemical reactions to create these processes, although, as CSO Stanisław Jastrzębski explained, they do it in reverse.
“Synthesis planning can be called a game,” he said. “With each movement in this game, instead of moving a piece on a board, we break a chemical bond between a pair of atoms. The object of the game is to break down a target molecule into molecules which can then be bought in the market and used to synthesize the target. We use algorithms similar to those used by DeepMind to master Go or chess to find the synthetic route. “
Co-founders Piotr Byrski and Paweł Włodarczyk-Pruszyński note that predicting organic reactions is not child’s play and that they have devoted a lot of resources to making their system efficient and verifiable. The theoretical pathways they produce seem plausible but have yet to be tested, which they regularly do internally so that companies see that Molecule.one only sells good ideas but achievable ideas.
Since starting out at Disrupt, the company has acquired a number of customers with annual contracts, Byrski said, and has deployed a number of features on the platform. Włodarczyk-Pruszyński said their effectiveness has also increased.
Image credits: Molecule.a
“Our system has matured and we have extended our platform to support scheduling the synthesis of thousands of molecules per hour,” he said. “This functionality is incredibly useful when combined with AI systems for drug discovery that generate large numbers of candidate molecules. All of these improvements have helped us gain confidence in the industry and initiate collaborations with relevant parties.
The problem certainly becomes a scaling problem when your clients start asking questions about the pathways for hundreds of thousands of possible therapeutic molecules rather than a handful. For them, if they have to bear the cost of manufacturing, it is worth paying up front to see if any of the compounds they are studying are considerably easier to manufacture than another with a similar effect. Without simulating the whole process, it’s hard to say for sure, so they can just send the list to Molecule.one and retrieve the report a few days later.
Image credits: Molecule.one
The team can’t share any of their client’s successes (although there probably has been some) because of course all of this work is highly confidential. But they said that, like many biotech companies, they were doing what they could to support COVID-related therapies.
“We is part of our platform for free to eligible researchers working on COVID drug discovery. This has resulted in a lasting collaboration with the LambdaZero project at MILA, who is advised by Professor Yoshua Bengio, ”Byrski said.
This also offered the opportunity to test their new scaling methods, because for such a project many candidate molecules must be evaluated, not only for their efficacy, but also for their ability to be easily manufactured.
“We are incredibly excited about this area in general as it allows crossing new regions of chemical space, which offers enormous promise in terms of finding drugs that have yet to be synthesized,” Byrski said. .
The roundtable was led by Atmos Ventures, with a long list of participating investors: AME Cloud Ventures, Cherubic Ventures, Firlej Kastory, Inventures, Luminous Ventures, Sunfish Partners and individuals including Bayer CEO Sebastian Guth.
The company plans to use the money to expand the team and continue to grow in general; it also plans to open new offices in the United States and Western Europe (they are based in Poland).
