Fig. 4

WGCNA identifies modules that are differentially associated with pathology and the AD and resilient groups. A Module-trait correlation of the 26 modules identified in relation to group and AD pathology with their relatedness. Notably, some modules correlate with pathology (ABC score for Thal, Braak and CERAD stages) and were significantly different between the different groups (such as the black, tan or saddlebrown modules), which were chosen as modules of interest. Significant module correlations with pathology and significant differences between groups are highlighted with p < 0.05: *, p < 0.01: ** and p < 0.001: ***. Trends between the different groups are indicated with p-values. B Results of the gene ontology (GO) and pathway analyses related to the genes in each module and the hub genes which could be regulators of these modules. GO and pathways of modules which correlate significantly with resilient or control are in boxes with solid lines and other relevant modules with dashed lines. C Heatmap of relative VST-transformed counts across samples of the genes in selected modules of interest. Note that donors cluster as group based on gene-expression in these selected modules. D Top GSEA results between the resilient and AD groups using the modules identified by WGCNA as input. Modules significantly enriched (padj < 0.05) are plotted with normalized enrichment score. E Quadrant plot of normalized enrichment score (NES) of modules in bulk RNA-seq. gene expression using GSEA. Both the resilient group and AD group are compared and plotted against controls. Modules significantly enriched (padj < 0.05) in resilient versus control are depicted in blue, AD versus control in red and in both groups versus control in yellow. The pink, midnight blue, steelblue and skyblue show an opposite enrichment, highlighting additional biologically relevant modules