Table 1.
Silenced TFs in cancer undergo preferential promoter hypermethylation in comparison with all cancer underexpressed genes
nTF | nU | P(DNAm) | P(CNV) | P(Mutation) | |
---|---|---|---|---|---|
LSCC | 95 | 7697 | 0.0005 | 0.888 | 0.898 |
LUAD | 94 | 6673 | 0.00004 | 0.926 | 0.032 |
KIRC | 68 | 7814 | 0.055 | 0.887 | 0.227 |
KIRP | 82 | 7465 | 0.039 | 0.998 | 0.555 |
BLCA | 81 | 5097 | 0.098 | 0.191 | 0.915 |
COAD | 98 | 7934 | 0.004 | 0.987 | 0.876 |
Combined Fisher test | <1e-8 | 0.984 | 0.471 |
The first six rows label the TCGA cancer type. Columns label the number of TFs expressed in the normal tissue and underexpressed in cancer (nTF), the number of all genes underexpressed in cancer (nU), the P value from a Wilcoxon rank sum test assessing whether promoter hypermethylation at the TFs is more frequent than for all underexpressed genes with DNAm information (DNAm), the P value from a Wilcoxon rank sum test assessing whether CNV loss of the TFs is more frequent than for all underexpressed genes with CNV information (CNV) and the P value from a Wilcoxon rank sum test assessing whether inactivating mutation of the TFs is more frequent than for all underexpressed genes with mutational information (Mutation). The last row lists P values of a meta-analysis over all cancer types using a combined Fisher test