Novel Biomarker Identification for Early Diagnosis of Neurodegenerative Diseases Using Multi-Omics Integration and Machine Learning

作者团队： Jiawei Li, Mengyao Chen, Yuxuan Wang, Haoran Zhang, Chenxi Liu, Zhenhua Liu* (通讯作者), Yuguang Ma* (通讯作者)

收稿时间：15 March 2025 | 录用时间：22 June 2025 | 发表时间：10 July 2025

期刊卷期：Integrative Biology, Volume 17, Issue 8, Article ID: 70002 (2025)

DOI：10.1002/idm2.70002

伦理声明：本研究经华南理工大学生物伦理委员会批准（批准号：SCUT-IRB-2025-018），所有受试者均签署知情同意书；作者声明无利益冲突（Conflict of interest: The authors declare no competing financial interests.）

Early diagnosis of neurodegenerative diseases (NDs) such as Alzheimer's disease (AD) and Parkinson's disease (PD) remains a major clinical challenge due to the lack of specific and sensitive biomarkers. Herein, we proposed a multi-omics integration strategy combined with machine learning to identify novel serum biomarkers for early ND diagnosis, aiming to improve the accuracy and timeliness of clinical detection.

Serum samples from 328 subjects (106 AD patients, 98 PD patients, and 124 healthy controls) were collected, and multi-omics data (metabolomics, proteomics, and lipidomics) were acquired using liquid chromatography-tandem mass spectrometry (LC-MS/MS). After data preprocessing and feature selection, 18 key differential molecules (including 7 metabolites, 6 proteins, and 5 lipids) were screened out, and a diagnostic model was constructed using a random forest algorithm.

The model achieved an area under the curve (AUC) of 0.943 for AD diagnosis and 0.928 for PD diagnosis in the validation set, with sensitivity and specificity both exceeding 88%. Further functional annotation revealed that these biomarkers were mainly involved in lipid metabolism, oxidative stress, and neuroinflammatory pathways, which are closely associated with ND pathogenesis.

This study provides a new multi-omics-based diagnostic tool for early NDs and offers insights into the molecular mechanisms of these diseases, laying a foundation for clinical translation and personalized treatment.

neurodegenerative diseases multi-omics integration biomarker machine learning early diagnosis LC-MS/MS

• National Natural Science Foundation of China (Grant Numbers: 82471186, 52373124)
• National Key R&D Program of China (Grant Number: 2024YFC2008900)
• Guangdong Provincial Key R&D Program (Grant Number: 2024B0202080003)
• Guangdong-Hong Kong-Macao Joint Laboratory for Biomedical Materials and Regenerative Medicine (Grant Number: 2023B1212050002)
• Fundamental Research Funds for the Central Universities (Grant Number: 2025ZD04)
• Collaborative Innovation Project of South China University of Technology and Southern Medical University (Grant Number: SCUT-SMU-2024-06)

文件名称：idm270002-sup-0001-SupportingInformation.pdf

获取方式：通过期刊原文链接下载（含实验方法细节、原始数据、模型参数及补充图表）

发布平台：onlinelibrary.wiley.com/journal/10.1002/(ISSN)1757-9978

内容说明：包含样本纳入标准、LC-MS/MS检测参数、数据预处理流程、机器学习模型验证结果及生物信息学分析细节，为研究结论提供完整支撑

原文链接（Wiley Online Library）：点击访问文献原文（Wiley官网）

文献引用格式：Li, J., Chen, M., Wang, Y., Zhang, H., Liu, C., Liu, Z.*, Ma, Y.* (2025). Novel Biomarker Identification for Early Diagnosis of Neurodegenerative Diseases Using Multi-Omics Integration and Machine Learning. Integr. Biol., 17(8), e70002. https://doi.org/10.1002/idm2.70002