Comparative Analysis of search Approaches to Discover Donor Molecules for Organic Solar Cells

TL;DR

This paper compares search algorithms for discovering donor molecules in organic solar cells, demonstrating that Bayesian optimization significantly outperforms random search in large chemical spaces.

Contribution

The study introduces stk-search, a Python package for exploring chemical space, and provides a comparative analysis of search algorithms, highlighting Bayesian optimization's superior performance in large-scale molecule discovery.

Findings

Bayesian optimization finds 1000 times more promising molecules than random search.

Performance differences between algorithms are minimal in small spaces but substantial in large spaces.

The Python package stk-search enables efficient exploration of vast chemical spaces.

Abstract

Identifying organic molecules with desirable properties from the extensive chemical space can be challenging, particularly when property evaluation methods are time-consuming and resource intensive. In this study, we illustrate this challenge by exploring the chemical space of large oligomers, constructed from monomeric building blocks, for potential use in organic photovoltaics (OPV). For this purpose, we developed a python package to search the chemical space using a building block approach: stk-search. We use stk-search to compare a variety of search algorithms, including those based upon Bayesian optimization and evolutionary approaches. Initially, we evaluated and compared the performance of different search algorithms within a precomputed search space. We then extended our investigation to the vast chemical space of molecules formed of 6 building blocks (6-mers), comprising over $10^{14}$ molecules. Notably, while some algorithms show only marginal improvements over a random search approach in a relatively small, precomputed, search space, their performance in the larger chemical space is orders of magnitude better. Specifically, Bayesian optimization identified a thousand times more promising molecules with the desired properties compared to random search, using the same computational resources.