Fig. 2: Multi-omics indicates ER alterations to microglia and the extracellular matrix.

A Results from Experiment 2 (n = 6/group) (Supplementary Table 4). The first analysis utilized GSEA to conduct full transcriptome pathway analysis. Significant pathways (p < 0.05) were condensed based on semantic similarity using Pathway Hunter (small left labels, and these categories were further condensed into functional themes based on a priori knowledge (leftmost labels)). Enrichment scores for individual pathways are represented by the heatmap (with EEvSH on the left, ERvSH in the middle, and ERvEE on the right), and the degree of each category’s enrichment is represented by the p value dots on the right. The enrichment scores show how much (magnitude) and in what direction (yellow is up, blue is down) a theme was changed by the model, while the p values show the contribution of each category to these changes. B The second analysis utilized Enrichr to conduct targeted pathway analysis and generated a similar set of results. Themes that replicated between these two pathway analyses are shown in the replicated pathways heatmap, with darker green indicating more involvement. C The third analysis utilized iLINCS to look for compounds that would replicate (concordant) or reverse (discordant) the present phenotypes. D We also used the leading-edge genes from GSEA to conduct cell type analysis. Collectively, these pipeline results start to implicate microglia and the extracellular matrix (ECM) in ER. E Similar analyses were run on the data from Experiment 3 (n = 10/group pooled). The RNAseq results are presented in Supplementary Fig. 5, while the proteomics and kinomics results are presented in Supplementary Fig. 6. Those results are summarized and compared to the above results here, and they provide more support for microglia and the ECM. This replication across cohorts, omics platforms, and analyses gives us higher confidence in these targets’ involvement. F Hypothesis-driven analyses of microglia and ECM-related genes further implicates both at multiple levels and points to a pattern of upregulation in EE (yellow), followed by downregulation in ER (blue).