Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis

BackgroundThe “gut–skin axis” has been proposed to play an important role in the development and symptoms of atopic dermatitis. Therefore, we have constructed an interpretable machine learning framework to quantitatively screen key gut flora.MethodsThe 16S rRNA dataset, after applying the centered l...

Full description

Saved in:
Bibliographic Details
Main Authors: Jingtai Ma, Yiting Fang, Shiqi Li, Lilian Zeng, Siyi Chen, Zhifeng Li, Guiyuan Ji, Xingfen Yang, Wei Wu
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-05-01
Series:Frontiers in Immunology
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fimmu.2025.1528046/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850282531907174400
author Jingtai Ma
Jingtai Ma
Yiting Fang
Yiting Fang
Shiqi Li
Shiqi Li
Lilian Zeng
Siyi Chen
Zhifeng Li
Guiyuan Ji
Xingfen Yang
Wei Wu
Wei Wu
author_facet Jingtai Ma
Jingtai Ma
Yiting Fang
Yiting Fang
Shiqi Li
Shiqi Li
Lilian Zeng
Siyi Chen
Zhifeng Li
Guiyuan Ji
Xingfen Yang
Wei Wu
Wei Wu
author_sort Jingtai Ma
collection DOAJ
description BackgroundThe “gut–skin axis” has been proposed to play an important role in the development and symptoms of atopic dermatitis. Therefore, we have constructed an interpretable machine learning framework to quantitatively screen key gut flora.MethodsThe 16S rRNA dataset, after applying the centered log-ratio transformation, was analyzed using five different machine learning models: random forest, light gradient boosting machine, extreme gradient boosting, support vector machine with radial kernel, and logistic regression. Interpretable machine learning methods, such as SHAP values, were used to identify significant features associated with atopic dermatitis.ResultsRandom forest performed better than the other “tree” models in the validation partitions. The SHAP global dependency plot indicated that Bifidobacterium ranked as the strongest predictive factor across all prediction horizons, although the SHAP values for some features were still higher in support vector machine and logistic regression models. The SHAP partial dependency plot for “tree” models showed that the best segmentation point for Bifidobacterium was further from the origin compared to other features in the respective models, quantitatively reflecting differences in gut microbiota.ConclusionMachine learning models combined with SHAP could be used to quantitatively screen key gut flora in atopic dermatitis patients, providing doctors with an intuitive understanding of 16S rRNA sequencing data to support precision medicine in care and recovery.
format Article
id doaj-art-ab27d4d5fa044e26a0c63bc09f87359c
institution OA Journals
issn 1664-3224
language English
publishDate 2025-05-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Immunology
spelling doaj-art-ab27d4d5fa044e26a0c63bc09f87359c2025-08-20T01:47:58ZengFrontiers Media S.A.Frontiers in Immunology1664-32242025-05-011610.3389/fimmu.2025.15280461528046Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitisJingtai Ma0Jingtai Ma1Yiting Fang2Yiting Fang3Shiqi Li4Shiqi Li5Lilian Zeng6Siyi Chen7Zhifeng Li8Guiyuan Ji9Xingfen Yang10Wei Wu11Wei Wu12National Medical Products Administration (NMPA) Key Laboratory for Safety Evaluation of Cosmetics, Guangdong Provincial Key Laboratory of Tropical Disease Research, School of Public Health, Southern Medical University, Guangzhou, ChinaGuangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaNational Medical Products Administration (NMPA) Key Laboratory for Safety Evaluation of Cosmetics, Guangdong Provincial Key Laboratory of Tropical Disease Research, School of Public Health, Southern Medical University, Guangzhou, ChinaGuangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaNational Medical Products Administration (NMPA) Key Laboratory for Safety Evaluation of Cosmetics, Guangdong Provincial Key Laboratory of Tropical Disease Research, School of Public Health, Southern Medical University, Guangzhou, ChinaGuangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaGuangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaGuangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaGuangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaGuangdong Provincial Institute of Public Health, Guangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaNational Medical Products Administration (NMPA) Key Laboratory for Safety Evaluation of Cosmetics, Guangdong Provincial Key Laboratory of Tropical Disease Research, School of Public Health, Southern Medical University, Guangzhou, ChinaNational Medical Products Administration (NMPA) Key Laboratory for Safety Evaluation of Cosmetics, Guangdong Provincial Key Laboratory of Tropical Disease Research, School of Public Health, Southern Medical University, Guangzhou, ChinaGuangdong Provincial Center for Disease Control and Prevention, Guangzhou, ChinaBackgroundThe “gut–skin axis” has been proposed to play an important role in the development and symptoms of atopic dermatitis. Therefore, we have constructed an interpretable machine learning framework to quantitatively screen key gut flora.MethodsThe 16S rRNA dataset, after applying the centered log-ratio transformation, was analyzed using five different machine learning models: random forest, light gradient boosting machine, extreme gradient boosting, support vector machine with radial kernel, and logistic regression. Interpretable machine learning methods, such as SHAP values, were used to identify significant features associated with atopic dermatitis.ResultsRandom forest performed better than the other “tree” models in the validation partitions. The SHAP global dependency plot indicated that Bifidobacterium ranked as the strongest predictive factor across all prediction horizons, although the SHAP values for some features were still higher in support vector machine and logistic regression models. The SHAP partial dependency plot for “tree” models showed that the best segmentation point for Bifidobacterium was further from the origin compared to other features in the respective models, quantitatively reflecting differences in gut microbiota.ConclusionMachine learning models combined with SHAP could be used to quantitatively screen key gut flora in atopic dermatitis patients, providing doctors with an intuitive understanding of 16S rRNA sequencing data to support precision medicine in care and recovery.https://www.frontiersin.org/articles/10.3389/fimmu.2025.1528046/fullmachine learningrandom forestlight gradient boosting machineextreme gradient boostingSHAP valuepartial dependence plot
spellingShingle Jingtai Ma
Jingtai Ma
Yiting Fang
Yiting Fang
Shiqi Li
Shiqi Li
Lilian Zeng
Siyi Chen
Zhifeng Li
Guiyuan Ji
Xingfen Yang
Wei Wu
Wei Wu
Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis
Frontiers in Immunology
machine learning
random forest
light gradient boosting machine
extreme gradient boosting
SHAP value
partial dependence plot
title Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis
title_full Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis
title_fullStr Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis
title_full_unstemmed Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis
title_short Interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis
title_sort interpretable machine learning algorithms reveal gut microbiome features associated with atopic dermatitis
topic machine learning
random forest
light gradient boosting machine
extreme gradient boosting
SHAP value
partial dependence plot
url https://www.frontiersin.org/articles/10.3389/fimmu.2025.1528046/full
work_keys_str_mv AT jingtaima interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT jingtaima interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT yitingfang interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT yitingfang interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT shiqili interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT shiqili interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT lilianzeng interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT siyichen interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT zhifengli interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT guiyuanji interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT xingfenyang interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT weiwu interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis
AT weiwu interpretablemachinelearningalgorithmsrevealgutmicrobiomefeaturesassociatedwithatopicdermatitis