Multimodal protective and susceptibility clusters in paediatric atopic dermatitis: A machine learning-based, data-driven observational study
11 affiliations
- 1Swiss Institute of Allergy and Asthma Research (SIAF), University of Zurich, Davos, Switzerland
- 2Swiss Institute of Bioinformatics (SIB), Lausanne, Switzerland
- 3Division of Dermatology, University of Cape Town, Cape Town, South Africa
- 4APC Microbiome Ireland, University College Cork, Cork, Ireland
- 5University of Applied Sciences of the Grisons, Chur, Switzerland
- 6ETHZ, D-BSSE, Basel, Switzerland
- 7Christine Kühne-Center for Allergy Research and Education, Davos, Switzerland
- 8Department of Environmental Health, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, United States of America
- 9Division of Paediatric Allergy, Department of Paediatrics and Child Health, University of Cape Town, Cape Town, South Africa
- 10Department of Medicine, University College Cork, Cork, Ireland
- 11School of Microbiology, University College Cork, Cork, Ireland
Abstract
Background: Atopic dermatitis (AD) is a chronic inflammatory skin disease that typically develops in early childhood. Differences in AD prevalence and allergy sensitisation patterns have been observed in African populations, including the AmaXhosa population in South Africa, suggesting alternative pathogenic and immune mechanisms underlying AD. Differences in AD prevalence have also been documented between urban and rural communities, making AmaXhosa children, who share a common ethnogenetic background but differ in environmental exposures, a unique population in which to investigate environmental and immune mechanisms underlying AD. To address this, we performed a machine learning (ML)-based multimodal observational study to identify features associated with AD in 217 AmaXhosa children. Methods and findings: To gain deeper insights into AD pathogenesis, we re-analysed a previously established multimodal dataset comprising environmental, cytokine, antibody, and transcriptomic data from healthy AmaXhosa children and children with AD, aged 12–36 months, living in rural or urban settings. We applied ML to analyse each data modality individually and subsequently integrate them to identify multimodal signatures associated with AD. Specifically, we used the GeneSelectR workflow to identify informative genes, SHAP values to explain the ML outputs, and DIABLO to integrate the datasets and identify protective and susceptibility clusters. Conclusions: Complementary ML approaches enabled explainable analysis of a complex multimodal dataset. Integrated analyses identified a multimodal protective cluster comprising environmental, cytokine, and transcriptomic features, together with two AD susceptibility clusters, one dominated by transcriptomic features and the other by correlated cytokine and antibody levels. These findings improve our understanding of the factors associated with AD. Furthermore, the feature-selection and integration workflow provides a framework for analysing complex multimodal datasets in future biomedical research. Why was this study done?: What did the researchers do and find?: What do these findings mean?: In this study, Damir Zhakparov and colleagues use machine learning to identify multimodal feature clusters associated with paediatric atopic dermatitis in AmaXhosa children in South Africa, highlighting a rural-environment-linked cluster associated with healthy skin and two susceptibility-related clusters.
Related
- AD-REAL: A Prospective, Multinational, Observational Cohort Study of Clinical Effectiveness and Treatment Discontinuation of Oral Systemics for Moderate-to-Severe Atopic Dermatitis
- A novel MSMO1 variant presenting with severe psoriasiform dermatitis: A case from Saudi Arabia
- Isoleucine and asparagine influence the growth of staphylococcal strains in atopic dermatitis
- Comment on "The utility of magnetic resonance imaging of the spinal cord in paediatric transverse myelitis: A systematic review
- Multiplex molecular respiratory diagnostic panels and antimicrobial use in paediatric respiratory infections: a systematic review and meta-analysis protocol.
BibTeX
@article{zhakparov2026multimodal,
title = {{Multimodal protective and susceptibility clusters in paediatric atopic dermatitis: A machine learning-based, data-driven observational study}},
author = {Damir Zhakparov and Nonhlanhla Lunjani and Marco Schmid and Kathleen Moriarty and Damian Roquero and Anita Dreher and Jeannette I. Heldstab-Kast and Kari C. Nadeau and Cezmi Akdis and Michael Levin and Carol Hlela and Milena Sokolowska and Liam O’Mahony and Katja Baerenfaller},
journal = {PLOS Medicine},
year = {2026},
month = sep,
doi = {10.1371/journal.pmed.1004917},
url = {https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004917},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Multimodal protective and susceptibility clusters in paediatric atopic dermatitis: A machine learning-based, data-driven observational study
AU - Damir Zhakparov
AU - Nonhlanhla Lunjani
AU - Marco Schmid
AU - Kathleen Moriarty
AU - Damian Roquero
AU - Anita Dreher
AU - Jeannette I. Heldstab-Kast
AU - Kari C. Nadeau
AU - Cezmi Akdis
AU - Michael Levin
AU - Carol Hlela
AU - Milena Sokolowska
AU - Liam O’Mahony
AU - Katja Baerenfaller
PY - 2026
DA - 2026/09/08
JO - PLOS Medicine
DO - 10.1371/journal.pmed.1004917
UR - https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004917
AB - Background: Atopic dermatitis (AD) is a chronic inflammatory skin disease that typically develops in early childhood. Differences in AD prevalence and allergy sensitisation patterns have been observed in African populations, including the AmaXhosa population in South Africa, suggesting alternative pathogenic and immune mechanisms underlying AD. Differences in AD prevalence have also been documented between urban and rural communities, making AmaXhosa children, who share a common ethnogenetic background but differ in environmental exposures, a unique population in which to investigate environmental and immune mechanisms underlying AD. To address this, we performed a machine learning (ML)-based multimodal observational study to identify features associated with AD in 217 AmaXhosa children. Methods and findings: To gain deeper insights into AD pathogenesis, we re-analysed a previously established multimodal dataset comprising environmental, cytokine, antibody, and transcriptomic data from healthy AmaXhosa children and children with AD, aged 12–36 months, living in rural or urban settings. We applied ML to analyse each data modality individually and subsequently integrate them to identify multimodal signatures associated with AD. Specifically, we used the GeneSelectR workflow to identify informative genes, SHAP values to explain the ML outputs, and DIABLO to integrate the datasets and identify protective and susceptibility clusters. Conclusions: Complementary ML approaches enabled explainable analysis of a complex multimodal dataset. Integrated analyses identified a multimodal protective cluster comprising environmental, cytokine, and transcriptomic features, together with two AD susceptibility clusters, one dominated by transcriptomic features and the other by correlated cytokine and antibody levels. These findings improve our understanding of the factors associated with AD. Furthermore, the feature-selection and integration workflow provides a framework for analysing complex multimodal datasets in future biomedical research. Why was this study done?: What did the researchers do and find?: What do these findings mean?: In this study, Damir Zhakparov and colleagues use machine learning to identify multimodal feature clusters associated with paediatric atopic dermatitis in AmaXhosa children in South Africa, highlighting a rural-environment-linked cluster associated with healthy skin and two susceptibility-related clusters.
SN - 1549-1676
ER -