Figure 1

TIN-estimates are associated with the expression levels of splicing factor genes in seven cancer datasets. (a) In seven of the cancer datasets (plots with grey background) there were strong and significant correlations (horizontal axes) between TIN-estimates and the expression levels of ≥41% of the totally 280 splicing factor genes. In the other three datasets, the correlations were mainly non-significant. Yellow and grey bars represent the percentages of splicing factor genes with significant and non-significant correlations, respectively. (b) In two-dimensional principal components analysis based on the expression levels of splicing factor genes (n = 280), samples were generally separated according to TIN-estimates (only samples with TIN-estimates ≥ ±1.0 were included, the number and percentages of samples are indicated for each dataset) in the same seven datasets (with grey background). In colorectal cancer series II and prostate cancer series I, the separation was not complete. In the gastric cancer dataset and lung cancer series I, there were few samples with TIN-estimates ≥ ±1.0. In the neuroblastoma dataset, there was no clear separation. PC, principal component.