IMR Press / FBL / Volume 28 / Issue 11 / DOI: 10.31083/j.fbl2811284
Open Access Original Research
Identification of Whole-Blood DNA Methylation Signatures and Rules Associated with COVID-19 Severity
Show Less
1 Department of Science and Technology, Binzhou Medical University Hospital, 256603 Binzhou, Shandong, China
2 School of Life Sciences, Shanghai University, 200444 Shanghai, China
3 Changping Laboratory, 102206 Beijing, China
4 Key Laboratory of Stem Cell Biology, Shanghai Jiao Tong University School of Medicine (SJTUSM) & Shanghai Institutes for Biological Sciences (SIBS), Chinese Academy of Sciences (CAS), 200031 Shanghai, China
5 College of Information Engineering, Shanghai Maritime University, 201306 Shanghai, China
6 Department of Computer Science, Guangdong AIB Polytechnic College, 510507 Guangzhou, Guangdong, China
7 Bio-Med Big Data Center, CAS Key Laboratory of Computational Biology, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, 200031 Shanghai, China
8 CAS Key Laboratory of Tissue Microenvironment and Tumor, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, 200031 Shanghai, China
*Correspondence: (Tao Huang); (Yu-Dong Cai)
These authors contributed equally.
Front. Biosci. (Landmark Ed) 2023, 28(11), 284;
Submitted: 24 April 2023 | Revised: 29 June 2023 | Accepted: 25 July 2023 | Published: 8 November 2023
Copyright: © 2023 The Author(s). Published by IMR Press.
This is an open access article under the CC BY 4.0 license.

Background: Different severities of coronavirus disease 2019 (COVID-19) cause different levels of respiratory symptoms and systemic inflammation. DNA methylation, a heritable epigenetic process, also shows differential changes in different severities of COVID-19. DNA methylation is involved in regulating the activity of various immune cells and influences immune pathways associated with viral infections. It may also be involved in regulating the expression of genes associated with the progression of COVID-19. Methods: In this study, a sophisticated machine-learning workflow was designed to analyze whole-blood DNA methylation data from COVID-19 patients with different severities versus healthy controls. We aimed to understand the role of DNA methylation in the development of COVID-19. The sample set contained 101 negative controls, 360 mildly infected individuals, and 113 severely infected individuals. Each sample involved 768,067 methylation sites. Three feature-ranking algorithms (least absolute shrinkage and selection operator (LASSO), light gradient-boosting machine (LightGBM), and Monte Carlo feature selection (MCFS)) were used to rank and filter out sites highly correlated with COVID-19. Based on the obtained ranking results, a high-performance classification model was constructed by combining the feature incremental approach with four classification algorithms (decision tree (DT), k-nearest neighbor (kNN), random forest (RF), and support vector machine (SVM)). Results: Some essential methylation sites and decision rules were obtained. Conclusions: The genes (IGSF6, CD38, and TLR2) of some essential methylation sites were confirmed to play important roles in the immune system.

COVID-19 severity
DNA methylation
machine learning
2022YFF1203202/National Key R&D Program of China
XDA26040304/Strate-gic Priority Research Program of Chinese Academy of Sciences
XDB38050200/Strate-gic Priority Research Program of Chinese Academy of Sciences
202002/Fund of the Key Laboratory of Tissue Microenvironment and Tumor of Chinese Academy of Sciences
ZR2022MC072/Shandong Provincial Natural Science Foundation
Fig. 1.
Back to top