Fig. 1

Phosphoproteome analysis of patients with RCC sensitive or resistant to sunitinib. a Overview of the data filtering steps applied in phosphosite and phosphopeptide analysis, including the effect of each filter on the total number. b Heatmap of the differentially detected phosphosites (n = 78) in sensitive and primary resistant patients, split by group. The heatmap is a concatenation of 3 heatmaps created with R package ComplexHeatmap. The first and third heatmaps were created with log10-transformed intensity values for phosphosites that were uniquely identified (“exclusive”) in the sensitive resp resistant patient group and had a data presence of at least 30%. The second heatmap was created with log10-transformed intensity values for significantly differential phosphosites (“non-exclusive”; p, 0.05, FC ≥ 2). This heatmap was clustered by columns but not by rows. Instead, rows were sorted by fold change and split by the sign of the fold change (down-regulated phosphosites in the upper part, up-regulated phosphosites in the lower part). Column splitting was at the first split of the column clustering dendrogram, and dendrogram plotting was set to FALSE. The column ordering in the resulting concatenated heatmap was determined by the middle heatmap. No imputation of data is performed. Euclidean distance and Ward’s linkage method were used. Black squares indicate non-identified phosphosites in this subgroup. Histology = histological subtype as determined by pathologist review; PFS = progression free survival in months; NE = not evaluable. c Volcano plot of for statistical comparison of differential class 1 phosphosites between the Sensitive and Resistant groups were generated in R with the ggplot2 package. The top 10 significant phosphosites for each group are indicated by labeling. Labels are given for the phosphosite, not the specific type of phosphopeptide in which it was detected. d Boxplots of differentially activated kinases based on INKA analysis. P-values by Mann–Whitney U-test. X-axis: 2 groups (primary resistant versus sensitive patients). Y-axis: INKA score of the kinase, based on kinase- and substrate-centric analyses. e PTM-SEA identified site-specific signatures of kinase activities and signaling pathways, overrepresented in each of the 2 groups. Phosphosites were ranked using the quantity -10 * sign(logFC) * log10(P-Value), where the P-value and logFC were calculated in a differential analysis by limma and used as inputs to the 20161013 version of ssGSEA2.0.R. The PTM-sets were defined in ptm.sig.db.all.flanking.human.v1.9.0.gmt. Significantly enriched signatures are presented in this figure (p < 0.05). X-axis represents the enrichment score (negative score = enriched in sensitive patients, positive score = enriched in resistant patients)