Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Predicting Molecular Geometry

Predicting Molecular Geometry

VSEPR Theory for Determination of Electron Pair Geometries

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Cracking the Capsid Code: A Computationally Feasible Approach for Investigating Virus-Excipient Interactions in Biologics Design.

Journal of chemical theory and computation·2026

Same author

Solvent-Inclusive ML/MM Simulations: Assessments of Structural, Dynamical, and Thermodynamic Accuracy.

Journal of chemical information and modeling·2025

Same author

Thermodynamic Basis of Sugar-Dependent Polymer Stabilization: Informing Biologic Formulation Design.

The journal of physical chemistry. B·2025

Same author

Crystal Nucleation Kinetics and Mechanism: Influence of Interaction Potential.

The journal of physical chemistry. B·2025

Same author

Machine Learning Reveals Amine Type in Polymer Micelles Determines mRNA Binding, <i>In Vitro, and In Vivo</i> Performance for Lung-Selective Delivery.

JACS Au·2025

Same author

LeaPP: Learning Pathways to Polymorphs through Machine Learning Analysis of Atomic Trajectories.

Journal of chemical theory and computation·2025

Search research articles

Related Experiment Video

Updated: Jan 3, 2026

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning

Published on: March 3, 2023

A generalized deep learning approach for local structure identification in molecular simulations.

Ryan S DeFever¹, Colin Targonski², Steven W Hall¹

¹Department of Chemical & Biomolecular Engineering , Clemson University , Clemson , SC 29634 , USA.

Chemical Science

|November 27, 2019

Summary

We developed a novel machine learning method using PointNets to identify local structures in molecular simulations. This approach bypasses complex feature engineering, achieving high accuracy in crystal structure identification across various systems.

More Related Videos

Structure-Based Simulation and Sampling of Transcription Factor Protein Movements along DNA from Atomic-Scale Stepping to Coarse-Grained Diffusion

Structure-Based Simulation and Sampling of Transcription Factor Protein Movements along DNA from Atomic-Scale Stepping to Coarse-Grained Diffusion

Published on: March 1, 2022

Investigating Protein Sequence-structure-dynamics Relationships with Bio3D-web

Investigating Protein Sequence-structure-dynamics Relationships with Bio3D-web

Published on: July 16, 2017

Related Experiment Videos

Last Updated: Jan 3, 2026

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning

Published on: March 3, 2023

Structure-Based Simulation and Sampling of Transcription Factor Protein Movements along DNA from Atomic-Scale Stepping to Coarse-Grained Diffusion

Structure-Based Simulation and Sampling of Transcription Factor Protein Movements along DNA from Atomic-Scale Stepping to Coarse-Grained Diffusion

Published on: March 1, 2022

Investigating Protein Sequence-structure-dynamics Relationships with Bio3D-web

Investigating Protein Sequence-structure-dynamics Relationships with Bio3D-web

Published on: July 16, 2017

Area of Science:

Computational chemistry
Materials science
Machine learning

Background:

Identifying local structure in molecular simulations is crucial for understanding material properties.
Traditional methods using order parameters require system-specific, often complex, feature engineering.

Purpose of the Study:

To develop a generic machine learning framework for identifying local structural environments in molecular simulations.
To eliminate the need for system-specific feature engineering in structure identification.

Main Methods:

Adapted a PointNet, a type of neural network from computer vision, to analyze raw atomic positions from simulations.
Applied the method to crystal structure identification in Lennard-Jones, water, and mesophase systems.

Main Results:

Achieved up to 99.5% accuracy in crystal structure identification.
Demonstrated applicability to heterogeneous nucleation and predicting crystal phases near interfaces.
Successfully identified surface hydrophobicity from water molecule configurations.

Conclusions:

The PointNet-based approach offers a versatile, feature-engineering-free framework for local structure identification in diverse molecular simulations.
This method shows broad applicability for analyzing complex structural phenomena in materials and interfaces.