Algorithms for self-driving labs

Théophile Gaudin
Doctor of Philosophy

Department of Department of Computer Science University of Toronto

This dissertation investigates how self-driving laboratories can be improved with the help of machine learning and data management techniques. The first part of the dissertation presents two methods, namely ORGANIZER and RouteScore, that can be implemented in existing self-driving labs. ORGANIZER is a software framework that enables equipment from multiple labs and locations to work seamlessly together, while RouteScore is a tool that quantifies the cost of synthetic routes, where both human and automated platforms can collaborate. The second part of the dissertation focuses on predicting the outcomes of chemical reactions. This would increase the available chemical space of the self-driving labs, making them more versatile. The chemical reaction were predicted using language models and a molecular representation called SMILES.

For more information about the used Chemspeed solutions:

ISYNTH

Contact us to learn more about this exciting publication:

https://www.chemspeed.com/contact-us/