Figure 1

Selection, characterization, and in silico validation of pathway reporter genes. (A) The selection workflow. Eclipses indicate data sources, green dots candidate genes, boxes the selected biological pathways, and triangles the selected biological processes. The pathway-gene association graph is bipartite and non-directional, while the transcriptional regulatory network consists exclusively of genes and is directional. All steps but panel customization is performed using the internal data infrastructure and software pipeline. (B) Histograms of reporter gene counts per pathway, and pathways per reporter gene, respectively. (C) Functional enrichment analysis results of reporter genes, randomly selected genes with high-confidence GO/BP annotations, randomly selected genes irrespective of annotations, and genes represent on a typical microarray (Illumina HT-6), respectively. Counts of GO BP terms falling in each Benjamini-Hochberg adjusted p-value bin are represented by bars (in case of reporter genes) or dots. The leftmost bin corresponding to adjusted p < 0.05: 1934 GO terms are significantly overrepresented by the pathway reporter genes using this threshold. (D) Density plot of number of diseases associated with differential expression of pathway reporter genes (red line), compared with associations with randomly selected genes (gray lines, repeated N = 100). (E) PageRank centrality of pathway reporter genes in the MetaBase transcriptional regulatory network compared with that of other human genes (background).