Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Structure-Activity Relationships and Drug Design01:28

Structure-Activity Relationships and Drug Design

1.2K
Drug design is a dynamic field that involves discovering and developing new medications based on specific biological targets. This process heavily relies on structure-activity relationships (SAR) and quantitative structure-activity relationships (QSAR) to guide the design and optimization of efficient drugs.
SAR studies the intricate relationship between a drug's chemical structure and biological activity. It focuses on understanding how modifications to a drug's structure can influence...
1.2K

You might also read

Related Articles

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

Sort by
Same author

Partially Reduced and Stabilized Phase-Changed MoS<sub>2</sub> Hybrids for Vapor Molecular Absorption at Defects in Edge Sites.

Small science·2026
Same author

Capillary-driven, superhydrophilic microfluidic retainer for continuous salivary glucose monitoring.

Lab on a chip·2025
Same author

Evaluating Machine Learning Interatomic Potentials for Accurate and Scalable Modeling of Organometallic Precursors.

ACS applied materials & interfaces·2025
Same author

ECG-GraphNet: Advanced arrhythmia classification based on graph convolutional networks.

Heart rhythm O2·2025
Same author

Cation-Disordered High-Entropy Garnet Structures as Solid-State Electrolytes for All-Solid-State Batteries: Machine Learning-Driven Discovery.

ACS applied materials & interfaces·2025
Same author

Matini-Net: Versatile Material Informatics Research Framework for Feature Engineering and Deep Neural Network Design.

Journal of chemical information and modeling·2024

Related Experiment Video

Updated: Oct 4, 2025

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
06:19

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

580

A Comparative Study of the Performance for Predicting Biodegradability Classification: The Quantitative

Myeonghun Lee1, Kyoungmin Min2

  • 1School of Systems Biomedical Science, Soongsil University, 369 Sangdo-ro, Dongjak-gu, Seoul 06978, Republic of Korea.

ACS Omega
|February 7, 2022
PubMed
Summary

Graph convolutional network (GCN) models offer a simpler and more stable alternative to quantitative structure-activity relationship (QSAR) models for predicting chemical biodegradability. GCNs demonstrate comparable performance without complex descriptors, enhancing computational chemistry predictions.

More Related Videos

Comparison of Scale in a Photosynthetic Reactor System for Algal Remediation of Wastewater
05:40

Comparison of Scale in a Photosynthetic Reactor System for Algal Remediation of Wastewater

Published on: March 6, 2017

9.2K
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
07:13

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model

Published on: April 18, 2025

229

Related Experiment Videos

Last Updated: Oct 4, 2025

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
06:19

Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction

Published on: August 16, 2024

580
Comparison of Scale in a Photosynthetic Reactor System for Algal Remediation of Wastewater
05:40

Comparison of Scale in a Photosynthetic Reactor System for Algal Remediation of Wastewater

Published on: March 6, 2017

9.2K
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
07:13

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model

Published on: April 18, 2025

229

Area of Science:

  • Computational chemistry
  • Environmental science
  • Machine learning in chemistry

Background:

  • Predicting chemical biodegradability is crucial for environmental safety and chemical development.
  • Quantitative structure-activity relationship (QSAR) models are established but complex to implement.
  • There is a need for more accessible and stable methods for biodegradability prediction.

Purpose of the Study:

  • To introduce and evaluate the graph convolutional network (GCN) method for predicting chemical biodegradability.
  • To compare the performance of GCN models against traditional QSAR models.
  • To assess the implementation simplicity and stability of GCNs for molecular property prediction.

Main Methods:

  • Trained prediction models using a biodegradability dataset.
  • Developed QSAR models utilizing Mordred descriptors and MACCS fingerprints.
  • Developed GCN models employing molecular graphs.
  • Validated model performance through rigorous testing with varied dataset splits.

Main Results:

  • GCN models proved more straightforward to implement and exhibited greater stability compared to QSAR models.
  • GCNs achieved comparable specificity and sensitivity to QSAR models without requiring specific molecular descriptors or fingerprints.
  • Performance validation across 100 random training/test set configurations confirmed GCN robustness.

Conclusions:

  • Graph convolutional network (GCN) models show significant promise as a replacement for conventional QSAR models in biodegradability prediction.
  • The straightforward implementation and stability of GCNs make them a valuable tool for computational chemistry.
  • GCNs offer a versatile approach applicable to predicting various molecular types and properties.