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

You might also read

Related Articles

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

Sort by
Same author

Dual Mechanisms of pH-Dependent Tea Astringency: Binding and Metabolism of Polyphenols Revealed by In Vivo and In Vitro Oral Processing.

Foods (Basel, Switzerland)·2026
Same author

The cholesterol-dependent cytolysin promotes <i>Streptococcus</i> systemic spread and induces arachidonic acid accumulation-mediated lethality in a murine intraperitoneal infection model.

Infection and immunity·2026
Same author

HyperDC: A Non-Uniform Hypergraph Framework for Dual- and Higher-Order Drug Combination Recommendation Across Diverse Complex Diseases.

Journal of chemical information and modeling·2026
Same author

Protective role of cyanidin-3-O-glucoside in 2-amino-3-methylimidazo[4,5-f]quinoline-induced gastrointestinal injury: focus on oxidative stress, inflammation, intestinal barrier and gut microbiota.

Journal of the science of food and agriculture·2026
Same author

Relationship between nursing notes sentiment scores and length of hospital stay in elderly knee osteoarthritis patients undergoing total knee arthroplasty: a retrospective observational study.

BMC nursing·2026
Same author

Dynamic Changes and Correlations of Physicochemical Parameters, Flavor Compounds and Microbial Communities During Soy Sauce Koji Production.

Foods (Basel, Switzerland)·2026

Related Experiment Video

Updated: Aug 9, 2025

Measuring Oral Fatty Acid Thresholds, Fat Perception, Fatty Food Liking, and Papillae Density in Humans
10:29

Measuring Oral Fatty Acid Thresholds, Fat Perception, Fatty Food Liking, and Papillae Density in Humans

Published on: June 4, 2014

20.3K

FAPD: An Astringency Threshold and Astringency Type Prediction Database for Flavonoid Compounds Based on Machine

Tianyang Guo1, Fei Pan1,2, Zhiyong Cui3

  • 1School of Food and Health, Beijing Technology and Business University, Beijing, 100048, China.

Journal of Agricultural and Food Chemistry
|February 24, 2023
PubMed
Summary

Machine learning (ML) accelerates the discovery of astringent compounds, which cause puckering sensations from flavonoids. A new database predicts astringency thresholds and types, offering a novel approach to understanding food component structure-flavor relationships.

Keywords:
astringency thresholdastringency typeflavonoid astringency prediction database (FAPD)flavonoid compoundsmachine learning (ML)

More Related Videos

Protocol for Data Collection and Analysis Applied to Automated Facial Expression Analysis Technology and Temporal Analysis for Sensory Evaluation
07:12

Protocol for Data Collection and Analysis Applied to Automated Facial Expression Analysis Technology and Temporal Analysis for Sensory Evaluation

Published on: August 26, 2016

9.5K
A New Straightforward Method for Lipophilicity logP Measurement using 19F NMR Spectroscopy
09:32

A New Straightforward Method for Lipophilicity logP Measurement using 19F NMR Spectroscopy

Published on: January 30, 2019

14.5K

Related Experiment Videos

Last Updated: Aug 9, 2025

Measuring Oral Fatty Acid Thresholds, Fat Perception, Fatty Food Liking, and Papillae Density in Humans
10:29

Measuring Oral Fatty Acid Thresholds, Fat Perception, Fatty Food Liking, and Papillae Density in Humans

Published on: June 4, 2014

20.3K
Protocol for Data Collection and Analysis Applied to Automated Facial Expression Analysis Technology and Temporal Analysis for Sensory Evaluation
07:12

Protocol for Data Collection and Analysis Applied to Automated Facial Expression Analysis Technology and Temporal Analysis for Sensory Evaluation

Published on: August 26, 2016

9.5K
A New Straightforward Method for Lipophilicity logP Measurement using 19F NMR Spectroscopy
09:32

A New Straightforward Method for Lipophilicity logP Measurement using 19F NMR Spectroscopy

Published on: January 30, 2019

14.5K

Area of Science:

  • Food Chemistry
  • Computational Chemistry
  • Sensory Science

Background:

  • Astringency, a key sensory property in foods, is primarily associated with flavonoid compounds.
  • Traditional methods for identifying astringent compounds are inefficient and costly.
  • Machine learning (ML) offers a promising avenue to accelerate the discovery and prediction of astringent properties.

Purpose of the Study:

  • To develop a machine learning-based database, the Flavonoid Astringency Prediction Database (FAPD), for predicting astringency in flavonoid compounds.
  • To establish and compare ML models for predicting both the threshold and type of astringency.
  • To provide a new computational paradigm for investigating the relationship between molecular structure and flavor properties of food components.

Main Methods:

  • Hierarchical clustering analysis of Molecular Fingerprint Similarities (MFSs) for flavonoid compounds.
  • Development and evaluation of regression models (GPR, SVR, RF, GBDT) for astringency threshold prediction, with Random Forest (RF) identified as the best model.
  • Development and evaluation of classification models (RF, GBDT, GNB, SVM, kNN, SGD) for astringency type prediction, with Stochastic Gradient Descent (SGD) identified as the best model.
  • Interpretation of the RF model using SHapley Additive exPlanations (SHAP) and verification of predictions using t-Distributed Stochastic Neighbor Embedding (t-SNE).

Main Results:

  • The Random Forest (RF) model demonstrated superior performance in predicting astringency thresholds.
  • The Stochastic Gradient Descent (SGD) model was identified as the most effective for classifying astringency types.
  • The Flavonoid Astringency Prediction Database (FAPD) was constructed, incorporating over 1200 natural flavonoid compounds with predicted astringency properties.
  • Model predictions were validated, confirming the utility of ML in this domain.

Conclusions:

  • Machine learning models, specifically RF and SGD, can accurately predict the astringency threshold and type of flavonoid compounds.
  • The FAPD provides a valuable resource for accelerating research into the structure-property relationships of food components.
  • This ML-driven approach establishes a new paradigm for investigating molecular mechanisms underlying food sensory attributes.