One Set to Rule Them All: How to Obtain General Chemical Conditions via Bayesian Optimization over Curried Functions

TL;DR

This paper explores how Bayesian optimization over curried functions can efficiently identify general chemical conditions that work across related reactions, reducing the need for re-optimization in experimental planning.

Contribution

It introduces a formulation of Bayesian optimization over curried functions for finding general chemical parameters, demonstrating its effectiveness on real-world data.

Findings

Simple myopic strategies perform comparably to complex ones.

Effective exploration of parameter and task space is crucial.

General optima can transfer to unseen reaction examples.

Abstract

General parameters are highly desirable in the natural sciences - e.g., chemical reaction conditions that enable high yields across a range of related transformations. This has a significant practical impact since those general parameters can be transferred to related tasks without the need for laborious and time-intensive re-optimization. While Bayesian optimization (BO) is widely applied to find optimal parameter sets for specific tasks, it has remained underused in experiment planning towards such general optima. In this work, we consider the real-world problem of condition optimization for chemical reactions to study how performing generality-oriented BO can accelerate the identification of general optima, and whether these optima also translate to unseen examples. This is achieved through a careful formulation of the problem as an optimization over curried functions, as well as systematic evaluations of generality-oriented strategies for optimization tasks on real-world experimental data. We find that for generality-oriented optimization, simple myopic optimization strategies that decouple parameter and task selection perform comparably to more complex ones, and that effective optimization is merely determined by an effective exploration of both parameter and task space.