2.1. The Construction of the Pipeline

The pipeline consisted of three parts: (1) Differential methylation analysis. Firstly, differential methylation analysis was performed on a trait to obtain differentially methylated positions and genes (DMPs/DMGs), which was also a common analysis strategy for epigenome-wide association studies (EWAS) [19,20]. (2) Calculating the correlation between gene methylation and expression. Since DNA methylation, as one of the important epigenetic regulatory mechanisms, has a stable regulatory relationship with gene expression in terms of tissue and cells [21,22], the correlation matrix between the methylation position and annotated gene was generated using the matched DNA methylation and gene expression data in the same patients. DMGs correlated with gene expression (DMG-exp) were selected for further analysis based on correlation and significance. (3) Single-cell level analysis. Differential expression analysis was performed in cell clusters in single-cell datasets to identify differentially expressed genes and their related cell types via comprehensive and accurate cell annotation (SingleR [23]). The genes obtained above were further annotated to specific cell types to obtain cell type-specific and non-cell-type-specific genes (Figure 1). It should be pointed out that while exploring a certain trait, the different omics data should be taken from the same tissue to ensure the validity of the conclusions.

(A): All datasets are from whole blood, whereas the scRNA-seq dataset contains mild and severe COVID-19 samples. (B): The pipeline contains three parts: differential methylation analysis, DNAm-mRNA correlation analysis, and single-cell RNA-seq data analysis.