Fig 1.
Visualization of the gene co-expression network modules and module relationships.
The co-expression network was built by considering all TNFa stimulated and unstimulated samples together. Hierarchical clustering dendogram of 14,019 genes expressed in HUVEC cell lines, along with colors representing module assignments. Genes that are not assigned to any module are colored grey. The heatmap shows the topological overlap matrix, and darker coloring indicates higher topological overlap.
Fig 2.
Adjacency heatmap showing the relationships among the module eigengenes.
Module eigengenes can be thought of as the weighted average gene expression of all genes in a module. For each pair of eigengenes EI, EJ, adjacency is calculated as (1 + cor(EI, EJ))/2. Red represents positively correlated modules and blue represents negatively correlated modules.
Table 1.
Description of gene modules identified by WGCNA and the relationship of genes in modules to TNFa.
Table 2.
DMRs and their relationship to genes and GH elements in WGCNA modules.
Fig 3.
Disease Ontology categories for the 136 diseases that showed overrepresentation of green module genes.
Boxes are colored according to the top-level disease categories and labels show the second-level categories. The size of each box is proportional to the number of disease terms in that category with significant overrepresentation of green module genes. Some disease terms belong to more than one category (e.g., multiple sclerosis is both a “nervous system disease” and an “immune system disease”), but each term is only represented once. For “disease of anatomical entity” terms, squares are shaded by the proportion of terms that represent autoimmune/inflammatory diseases (e.g., 3 of 16 “gastrointestinal system disease”terms are autoimmune/inflammatory, while 9 of 11 “immune system disease” terms are). The full list of disease terms, with (manually curated) autoimmune/inflammatory terms highlighted, is given in S3 File.