Related Experiment Video
Updated: Feb 2, 2026

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Automated calculation of thermal rate coefficients using ring polymer molecular dynamics and machine-learning
I S Novikov1, Y V Suleimanov, A V Shapeev
1Skolkovo Institute of Science and Technology, Skolkovo Innovation Center, Nobel St. 3, Moscow 143026, Russia. a.shapeev@skoltech.ru.
This study introduces an automated method for calculating reaction rates using ring polymer molecular dynamics (RPMD) and machine learning. The approach efficiently builds potential energy surfaces, enabling accurate predictions for gas phase chemical reactions.
Area of Science:
- Computational Chemistry
- Chemical Physics
- Materials Science
Background:
- Calculating thermal rate coefficients for gas phase reactions is crucial for understanding chemical processes.
- Traditional methods for determining potential energy surfaces (PESs) can be computationally intensive and time-consuming.
- Accurate PESs are essential for reliable simulations of chemical reaction dynamics.
Purpose of the Study:
- To develop a fully automated methodology for calculating thermal rate coefficients of gas phase chemical reactions.
- To combine ring polymer molecular dynamics (RPMD) with machine-learning interatomic potentials that learn on-the-fly.
- To create a general tool for calculating RPMD thermal rate coefficients for polyatomic gas phase reactions.
Main Methods:
- The methodology integrates RPMD simulations with machine learning (ML) potentials.
- Potential energy surfaces (PESs) are constructed automatically and gradually from scratch.
- Data points for ML model training are selected and accumulated during RPMD simulations.
- The approach avoids artifacts by ensuring the ML model reliably describes the PES explored by RPMD trajectories.
Main Results:
- The methodology was tested on two representative thermally activated chemical reactions.
- PESs were generated by fitting to fewer than 5000 automatically generated structures.
- Calculated RPMD rate coefficients showed deviations within the typical convergence error of RPMDrate compared to reference values.
- The approach demonstrated efficiency in generating accurate PESs for chemical reaction dynamics.
Conclusions:
- The developed methodology offers a robust and automated approach for calculating thermal rate coefficients.
- The combination of RPMD and active learning ML potentials provides a reliable and efficient tool.
- Future work will focus on applying this methodology to complex-formation reactions, aiming for a universal tool.
Related Concept Videos
Avoidance Learning and Learned Helplessness
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
Associative Learning
Classical conditioning, also known...
Purposive Learning
Observational Learning
Learning Disabilities
Dyslexia
Dyslexia is a...
Introduction to Learning
In contrast to learned behaviors, unlearned behaviors such as crying, sexual...

