Fig. 2: Identification of molecular differences between IBC and nIBC preclinical models.

a Volcano plot representing gene expression differences between IBC and nIBC preclinical models. The X-axis indicates the log2-transformed gene expression fold change in IBC relative to nIBC. The Y-axis represents the −log10-transformed p value. The horizontal dashed line represents a nominal P value threshold of 5%. Genes color-coded yellow and blue are overexpressed in IBC and nIBC respectively at a false discovery adjusted p value of 10%. The top 10 overexpressed genes in IBC and nIBC cell lines are labeled. b Heatmap representing the correlation structure of 22 co-expression modules identified using WGCNA. Pearson correlation coefficients resulting from pairwise comparisons of the eigengenes of the different co-expression modules are coded according to a blue-red color scheme reflecting correlation coefficients ranging from −1 to 1. Row and columns are labeled with the names of the co-expression modules and are ordered according to an unsupervised hierarchical cluster analysis. The three co-expression cluster groups are indicated in orange squares. c Dot plot representing the result of a gene set enrichment analysis (GSEA) obtained by comparing the gene-module membership scores for each co-expression cluster to the hallmark gene sets. The co-expression modules are listed along the X-axis and a different facet is created for each co-expression cluster. Enriched hallmarks per modules are indicated using a dot, the color and size of which vary with respectively the normalized enrichment score (i.e., blue = low; red = high) and the −log10-transformed p value (i.e., small = less significant; large = more significant) that result from the GSEA.