Optimizing Chemical Reactions with Deep Reinforcement Learning
By Zhenpeng Zhou, Xiaocheng Li, Richard N. Zare. ACS Central Science, 2017
395 citations recorded by the source corpus.
Deep reinforcement learning was employed to optimize chemical reactions. Our model iteratively records the results of a chemical reaction and chooses new experimental conditions to improve the reaction outcome. This model outperformed a state-of-the-art blackbox optimization algorithm by using 71% fewer steps on both simulations and real reactions. Furthermore, we introduced an efficient exploration strategy by drawing the reaction conditions from certain probability distributions, which resulted in an improvement on regret from 0.062 to 0.039 compared with a deterministic policy. Combining the efficient exploration policy with accelerated microdroplet reactions, optimal reaction conditions were determined in 30 min for the four reactions considered, and a better understanding of the factors that control microdroplet reactions was reached. Moreover, our model showed a better performance after training on reactions with similar or even dissimilar underlying mechanisms, which demonstrates its learning ability.
Subfields: Artificial Intelligence, Biomedical Engineering, Materials Chemistry
View Paper: https://www.semanticscholar.org/paper/4ab08b2f1193d770c241b41f5d9f1c841a3663d3
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Hsin-Yuan Huang
- John Preskill
- Richard Kueng
- Alán Aspuru-Guzik
- Christoph Kreisbeck
- Loïc M. Roch
- Patrick J. Coles
- M. Cerezo
In Biomedical Engineering
In Materials Chemistry
- Alán Aspuru-Guzik
- Christoph J. Brabec
- Richard Kueng
- Lauv Patel
- David Wayne Ussery
- John Preskill
- Hsin-Yuan Huang
- Xiuzhen Huang
Citations in the corpus
- Prediction of higher-selectivity catalysts by computer-driven workflow and machine learning (2019, neutral): “Computer-assisted synthetic planning (5, 6), the prediction of organic reaction outcomes (7, 8), assisted medicinal chemistry discovery (9, 10), catalyst design (11, 12), the use of continuous molecular representations for automatic generation of new chemical structures (13), materials discovery (14”
- Machine Learning Methods in Drug Discovery (2020, background): “Various studies have been conducted to optimize AI-based chemical synthesis and reaction routes [85–87].”
- Information-theoretic bounds on quantum advantage in machine learning (2021, background): “Such experiments might explore, for instance, the outcome of a reaction in quantum chemistry [15], ground-state properties of a novel molecule or material [11,16–21], or the behavior of neutral atoms in an analog quantum simulator [22–24].”
- Accelerating the discovery of materials for clean energy in the era of smart automation (2018, methodology): “Recently, an autonomous infrastructure for the optimization of chemical reaction conditions through the Deep Reaction Optimizer (DRO), an algorithm based on deep reinforcement learning, was reported 57 .”