A systematic atlas of chaperome deregulation topologies across the human cancer landscape
Fig 3
Preferential upregulation of ATP-dependent chaperones in group 1 cancers.
Analysis of differential cancer gene expression of chaperome functional subsets. A. Comparing upregulation and downregulation of gene expression (∆GSA) of chaperones (n = 88) and co-chaperones (n = 244) reveals general chaperome upregulation for the majority of cancers (Group 1), while a small group of Group 2 cancers do not follow this trend. Colour code indicates chaperone up-regulation of gene expression, axes represent chaperone downregulation, co-chaperone upregulation and down-regulation of gene expression. s = median silhouette width (k-means clustering). B. Box-and-whisker plots highlight fractions of differentially expressed genes in each chaperome subset for Group 1 (red) and Group 2 (blue) cancers separately (see A). Differentially expressed genes in each set were obtained by linear modelling (Limma package in R), considering genes with p value < 0.05 (Benjamini-Hochberg corrected). Box boundaries, 25% and 75% quartiles; middle horizontal line, median; whiskers, quartile boundaries for values beyond 1.5 times the interquartile range; small circle, outlier. C. Assessing differential expression of ATP-dependent (n = 50) vs. ATP-independent (n = 38) chaperones highlights preferential upregulation of ATP-dependent chaperones in Group 1 cancers. Group 2 cancers do not follow this trend. s = median silhouette width (k-means clustering). D. Box-and-whisker plots (as in B.) show fractions of differentially expressed genes in the two sets of ATP-dependent and ATP-independent chaperones, partitioned by Group 1 and Group 2 cancers (see C).