Accelerated Chemical Reaction Optimization using Multi-Task Learning

30 January 2023, Version 1
This content is a preprint and has not undergone peer review at the time of posting.


Functionalization of C–H bonds is a key challenge in medicinal chemistry, particularly for fragment-based drug discovery (FBDD) where such transformations need to be executed in the presence of polar functionality necessary for fragment-protein binding. New technologies such as high-throughput experimentation and self-optimization have the potential to revolutionize synthetic approaches to challenging target molecules in FBDD. Recent work has shown the effectiveness of Bayesian optimization (BO) for the self-optimization of chemical reactions, however, in all previous cases these algorithmic procedures have started with no prior information about the reaction of interest. In this work, we explore the use of multi-task Bayesian optimization (MTBO) in several in silico case studies by leveraging reaction data collected during related historical optimization campaigns to accelerate the optimization of new reactions - this was performed for Suzuki-Miyaura and C–N couplings. This methodology was then translated to real-world, medicinal chemistry applications in the yield optimization of several pharmaceutical intermediates using an autonomous flow-based reactor platform. The use of the MTBO algorithm was shown to be successful in determining optimal conditions (both continuous and categorical variables) of unseen experimental C–H activation reactions with differing substrates, demonstrating up to a 98 % cost reduction when compared to industry-standard process optimization techniques. Our findings highlight the effectiveness of the methodology as an enabling tool in medicinal chemistry workflows, where efficient utilization of precious starting materials is particularly important. This work represents a step-change in the utilization of previously obtained reaction data and machine learning with the ultimate goal of accelerated reaction optimization.


multi-task learning
c-h activation
machine learning


Comments are not moderated before they are posted, but they can be removed by the site moderators if they are found to be in contravention of our Commenting Policy [opens in a new tab] - please read this policy before you post. Comments should be used for scholarly discussion of the content in question. You can find more information about how to use the commenting feature here [opens in a new tab] .
This site is protected by reCAPTCHA and the Google Privacy Policy [opens in a new tab] and Terms of Service [opens in a new tab] apply.