Thermal half-lives of azobenzene derivatives: virtual screening based on intersystem crossing using a machine learning potential

Abstract

Molecular photoswitches are the foundation of light-activated drugs. A key photoswitch is azobenzene, which exhibits trans-cis isomerism in response to light. The thermal half-life of the cis isomer is of crucial importance, since it controls the duration of the light-induced biological effect. Here we introduce a computational tool for predicting the thermal half-lives of azobenzene derivatives. Our automated approach uses a fast and accurate machine learning potential trained on quantum chemistry data. Building on well-established earlier evidence, we argue that thermal isomerization proceeds through rotation mediated by intersystem crossing, and incorporate this mechanism into our automated workflow. We use our approach to predict the thermal half-lives of 19,000 azobenzene derivatives. We explore trends and tradeoffs between barriers and absorption wavelengths, and open-source our data and software to accelerate research in photopharmacology.

Solving the Azobenzene Entropy Puzzle: Direct Evidence for Multi-State Reactivity

“…This will hopefully assist in the ongoing rational design of such photoswitches, for example photopharmacophores with tunable thermal half-lives. Just before submission of this work, we became aware of a preprint by Axelrod et al 52 that also concluded, based on spin-flip TDDFT calculations and somewhat simplified rate expressions, that the type II rotation mechanism is the correct one for the parent azobenzene thermal isomerization, and they used this assumption in the virtual screening of 19000 azobenzene derivatives using machine-learning techniques.…”

Data-driven discovery of molecular photoswitches with multioutput Gaussian processes

“…Sources of future work include the curation of an experimental dataset of the thermal reversion barriers to improve the predictive capabilities of machine learning models. Such a dataset would complement recent advances in machine learning prediction of thermal reversion barriers using quantum chemical photoswitch datasets, 82 as well as machine learning approaches for accelerating the speed of quantum chemical simulations themselves. 83 A further point of interest would be an investigation into how synthetic chemists may use model uncertainty estimates in the decision process to screen molecules e.g.…”