Fig. 2
From: Robust inflammatory breast cancer gene signature using nonparametric random forest analysis
Independent validation of IBC gene signature and its gene ontology/pathway analysis. a, b Validation of post-treatment samples from GSE5847 dataset and pre-treatment core biopsies samples from GSE111477 dataset, respectively. IBC probability plot, PCA scatter plot and unsupervised hierarchical clustering heatmaps are represented similar to Fig. 1. c (i) Venn plots for G59 overlap with 5 previous IBC gene signatures (see Additional file 1: Supp. Methods). (ii) Table indicating the accuracy of the signatures in GSE45581 and GSE111477 datasets (See details in Additional file 1: Supp. Methods). d Kaplan–Meier plot log-rank test for G59-predicted IBC like versus non-IBC like samples in TCGA (see details in Additional file 1: Supp. Methods). The p-value, hazard ratio (HR) and the 95% confidence interval of ratio are indicated. e Pie chart indicating the proportion of gene types in the signature. ncRNA: non-coding RNA. f Clustergrams of top 10 cellular component and pathway analysis of the signature genes, with overlapping genes highlighted (see Additional file 1: Table S3 and S4 for complete list)