Fig. 5

Transcription factor (TF) and global targeting associations involving the essential metastasis signature. a Venn diagram of the overlapping genes involving four gene sets: genes with high expression in breast cancer metastasis versus paired primary (blue, based on BIG, using a relaxed p-value of 4759 by paired t-test); genes essential in at least 10% of breast cancer cell lines (light orange, using effect score < − 0.75); TF genes with bound targets enriched (p < 0.01, one-sided Fisher’s exact test) in the essential gene signature (green, TF associations by Encode [34] data); and genes for which siRNA knockdown results in under-expression of genes that are significantly enriched (p < 0.01, one-sided Fisher’s exact test) in the essential metastasis signature (red, using a gene expression dataset of human umbilical endothelial cells transfected with siRNAs for 400 different genes [35]). b For the seven genes overlapping among three of the four gene sets from part a, the corresponding patterns are represented involving differential expression in breast cancer metastasis, CRISPR-based gene effect scoring in breast cancer cell lines, enrichment of TF-bound genes by Encode data, and enrichment of siRNA targets. Gene set enrichment patterns (by one-sided Fisher’s exact test) are represented for both essential and nonessential metastasis gene sets. c For both essential and nonessential metastasis signature genes (top and bottom, respectively), the corresponding patterns are represented involving differential expression in metastases, CRISPR-based gene effect scoring, binding 2 kb upstream for selected TFs, and differential expression in response to siRNA knockdown (GSE27869 dataset [35] involving profiling of cells after knockdown of each of 400 genes)