Machine Learning Based Plasma Proteomic Analysis Identifies Disease Differentiating Biomarkers in Atopic Dermatitis

BACKGROUND: Inflammatory diseases consist of a diverse group of conditions characterized by immune responses leading to inflammation causing damage and abnormality in function. Atopic dermatitis (AD) and ulcerative colitis (UC) both exhibit aberrant Th2 cell activity, leading to epithelial barrier dysfunction in the skin and gut, respectively. Although both diseases share similar features, differences in their tissue-specific pathology, clinical manifestations, and immune profiles may provide valuable insights into the distinct mechanisms underlying each disease. METHODS: We used the high-throughput Olink 384 proximity extension assay platform to profile plasma proteins from 58 patients with AD or UC and 29 healthy subjects. Differentially expressed proteins (DEPs) were identified and machine-learning (ML) model using least absolute shrinkage and selection operator (LASSO) regression was conducted to select disease-differentiating biomarkers. Pearson correlation analysis was performed to identify protein modules associated with severity of AD. RESULTS: DEP analysis yielded 85 (35 unique) and 99 (49 unique) upregulated proteins in AD and UC, respectively, compared to healthy subjects. We integrated an ML model and selected 24 proteins to distinguish between the diseases, which accurately predicted disease-differentiating biomarkers based on distinctive proteomic signatures and disease severity. Using nested cross-validation approach, we achieved robust classification performance with a mean area under the receiver operating characteristic curve of 0.99 ± 0.01. Correlation analysis with disease severity identified upregulated ML-selected proteins such as CCL13, CCL26, CD70, CDON, LY6D and MMP1 in AD, and ITM2A and REG4 in UC. Among these identified proteins, we suggest CDON, CD70, and LY6D, which had highly correlated expression levels, as AD-specific biomarkers. CONCLUSION: Our study provides insight into the application of ML algorithms for disease diagnosis, and our model may be expanded to other disease contexts. Our results identify inflammatory protein biomarkers and suggest their relevance in predicting disease severity, monitoring clinical progression and future therapeutic strategies.

Authors

Institutions

Publication Details

Journal
Journal of Korean Medical Science
Published
2026-11-16
DOI
https://doi.org/10.3346/jkms.2026.41.e276
Primary Topic
Dermatology and Skin Diseases
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Machine Learning Based Plasma Proteomic Analysis Identifies Disease Differentiating Biomarkers in Atopic Dermatitis

Christine Suh-Yun Joh, Yoon Ji Bang, Soyoung Jeong, Minsik Oh et al.
3 links
Journal of Korean Medical Science
Dermatology and Skin Diseases
article

Machine Learning Based Plasma Proteomic Analysis Identifies Disease Differentiating Biomarkers in Atopic Dermatitis

Christine Suh-Yun Joh, Yoon Ji Bang, Soyoung Jeong, Minsik Oh, Ji Hwan Moon, Jong Wook Jung, Jaeyun Park, Jasper Roldan Go, Hyun Je Kim
article en

Abstract

BACKGROUND: Inflammatory diseases consist of a diverse group of conditions characterized by immune responses leading to inflammation causing damage and abnormality in function. Atopic dermatitis (AD) and ulcerative colitis (UC) both exhibit aberrant Th2 cell activity, leading to epithelial barrier dysfunction in the skin and gut, respectively. Although both diseases share similar features, differences in their tissue-specific pathology, clinical manifestations, and immune profiles may provide valuable insights into the distinct mechanisms underlying each disease. METHODS: We used the high-throughput Olink 384 proximity extension assay platform to profile plasma proteins from 58 patients with AD or UC and 29 healthy subjects. Differentially expressed proteins (DEPs) were identified and machine-learning (ML) model using least absolute shrinkage and selection operator (LASSO) regression was conducted to select disease-differentiating biomarkers. Pearson correlation analysis was performed to identify protein modules associated with severity of AD. RESULTS: DEP analysis yielded 85 (35 unique) and 99 (49 unique) upregulated proteins in AD and UC, respectively, compared to healthy subjects. We integrated an ML model and selected 24 proteins to distinguish between the diseases, which accurately predicted disease-differentiating biomarkers based on distinctive proteomic signatures and disease severity. Using nested cross-validation approach, we achieved robust classification performance with a mean area under the receiver operating characteristic curve of 0.99 ± 0.01. Correlation analysis with disease severity identified upregulated ML-selected proteins such as CCL13, CCL26, CD70, CDON, LY6D and MMP1 in AD, and ITM2A and REG4 in UC. Among these identified proteins, we suggest CDON, CD70, and LY6D, which had highly correlated expression levels, as AD-specific biomarkers. CONCLUSION: Our study provides insight into the application of ML algorithms for disease diagnosis, and our model may be expanded to other disease contexts. Our results identify inflammatory protein biomarkers and suggest their relevance in predicting disease severity, monitoring clinical progression and future therapeutic strategies.

Journal of Korean Medical ScienceVol. 41(44)
Seoul National University (KR), Samsung Medical Center (KR), Myongji University (KR)
National Research Foundation of Korea, Institute for Information and Communications Technology Promotion
Good health and well-being
Dermatology and Skin Diseases
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon