Fig. 4

Prediction of patient survival. a All TCGA Glioblastoma (GBM) patients are ranked by EGFR expression; the top half patients are assigned as high group and the lower half are assigned as low. Kaplan Meier (KM) survival plot is shown for two groups. b The survival analysis is done in the same way as A for EGFR NEST scores. c For each gene, we calculated a death risk Z-score by Cox-PH model from either gene expression or NEST score. We compared the Z-scores for oncogenes (Onco) and tumor suppressor genes (TSG) based on the annotation from Vogelstein et al. The bottom and top of the boxes are the 25th and 75th percentiles (interquartile range). Whiskers on the top and bottom represent the maximum and minimum data points within the range represented by 1.5 times the interquartile range. The distribution of Z-scores is compared by Wilcoxon rank-sum test and three stars represent P value <0.001. d For each TCGA cohort, the difference of risk Z-scores computed from NEST was tested by Wilcoxon rank-sum test. The rank-sum Z-scores are plotted against the death rate of each cancer type, with Spearman’s rank correlation as title