A cross-disciplinary team of MIT researchers have published details of an artificial intelligence (AI) guided robotic platform for flow synthesis of small molecule organic compounds. The paper appeared in the August 9, 2019 issue of Science.
While small organic molecules are essential for many disciplines including drug discovery, the identification and development of synthetic routes creates a bottleneck due to the need for time-consuming manual tasks and lengthy design-synthesize-test iterations. Despite the availability of laboratory automation, experimental synthesis platforms still require manual configuration by expert chemists.
In this publication, the authors describe the development and use of a platform that combines artificial intelligence-driven synthesis planning, flow chemistry and a robotically controlled experimental platform to minimize the need for human intervention in the synthesis process from ideation through to manufacturing. Synthetic routes are proposed through generalization of millions of published chemical reactions and validated in silico to maximize the likelihood of success. Additional implementation details are determined by expert chemists and recorded in recipe files, which are executed by a modular continuous-flow platform that is automatically reconfigured by a robotic arm to set up the required unit operations and carry out the reaction.
"One of the major challenges in automating small molecule synthesis is the diversity of organic reactions and the difficulty in finding compatible reaction conditions to support multistep synthesis. Additionally, creating a system capable of supporting the range of reaction conditions in terms of temperatures, pressures and chemical compatibilities poses a serious engineering challenge," said Justin Lummiss, PhD, a co-first author on the Science publication who conducted research as a Postdoctoral Fellow at MIT.
Lummiss and Dale Thomas, PhD (another co-first author of the Science paper) are leveraging an AI-guided approach to address similar, longstanding challenges with custom peptide synthesis and purification at Mytide Therapeutics. Mytide is using AI to enable end-to-end peptide manufacturing in a truly autonomous fashion, including optimized reaction planning, real-time process optimization for in-sequence process modifications and robotics.
Similar challenges exist in the automation of small molecule and peptide synthesis. We have made important strides in addressing significant engineering obstacles for small molecules and are now focused on new challenges with peptide synthesis and manufacturing".
Dr. Justin Lummiss, Senior Scientist at Mytide
Thomas and Lummiss' current work at Mytide Therapeutics focuses on the development of an integrated platform for peptide manufacturing and discovery of novel peptide therapeutics. The AI-guided, automated platform is designed to address challenges not only in synthesis but also in purification of custom peptides.
"The fully automated platform we are developing at Mytide simplifies the challenges of peptide synthesis through mastery of a single reaction - amide bond formation," noted Dr. Thomas, a Co-Founder of Mytide. "By reducing the synthetic diversity of conventional approaches to peptide synthesis, we have been able to address longstanding challenges in synthesis and purification, as well as reducing barriers of purity bottlenecks that presently limit current large-scale peptide manufacturing for drug products".
Thomas and Lummiss are excited about the future applications of their current work at Mytide and its ability to have a meaningful impact. Their advances in AI-guided automation and technology are driving Mytide's ultimate goal of accelerating the discovery and development of peptide therapeutics targeting a wide range of diseases, including cancer therapy, infectious disease, and inflammatory disorders.
Radnet’s DeepHealth and HOPPR forge partnership to advance AI in healthcare