Table 2.
Top 10 most significant triplets identified by unsupervised MethReg analysis of TCGA COADREAD samples
| Triplet data | Annotations | DNAm group × TF activity | TF–target associations | Tumor versus normal comparison P-values | Literature support | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Probe ID | Position | TF | Target gene | CpG–target distance | TF.role | DNAm.effect | Estimate | P-value | Adj. P-value | Correlation | P-value | CpG | TF | Target gene | TF | Target gene |
| Promoter analysis | ||||||||||||||||
| cg00328227 | chr1: 108661703 | NFATC2 | HENMT1 | −176 | Repressor | Enhancing | −0.663 | 0 | 0 | −0.033 | 5.27E−01 | 3.19E−08 | 2.53E−01 | 8.23E−08 | (48–49,51) | (52,53) |
| cg25083481 | chr11: 1034989 | SNAI1 | MUC6 | 1727 | Activator | Attenuating | −0.413 | 1.77E−04 | 0 | 0.117 | 2.52E−02 | 8.98E−03 | 4.99E−10 | 1.97E−02 | (64,66–67) | (86) |
| cg02816729 | chr6: 168442419 | TEAD3 | SMOC2 | 1267 | Repressor | Attenuating | 0.323 | 7.77E−05 | 0 | 0.07 | 1.83E−01 | 2.96E−02 | 3.47E−02 | 7.32E−01 | (54) | (56,87) |
| cg12751565 | chr1: 206946642 | NFATC2 | PIGR | −175 | Activator | Enhancing | 0.282 | 1.09E−09 | 0 | 0.315 | 8.67E−10 | 6.49E−03 | 2.53E−01 | 1.88E−11 | (48–49,51) | (88) |
| cg05503887 | chr11: 34620378 | MEIS1 | EHF | −713 | Activator | Attenuating | −0.268 | 7.24E−05 | 0 | 0.347 | 1.10E−11 | 6.17E−01 | 1.84E−08 | 1.16E−02 | (68) | (89) |
| cg13542964 | chr4: 10457043 | ETV4 | ZNF518B | 381 | Activator | Enhancing | 0.217 | 2.08E−04 | 0 | 0.023 | 6.59E−01 | 1.31E−06 | 1.32E−14 | 1.77E−02 | (70,74) | (90) |
| Distal analysis | ||||||||||||||||
| cg14043104 | chr13: 74075149 | ISL2 | KLF12 | −80 092 | Activator | Attenuating | −0.222 | 1.03E−04 | 0 | 0.232 | 7.60E−06 | 3.89E−02 | 3.96E−02 | 1.20E−03 | (69) | (91) |
| cg09217215 | chr11: 31808034 | ZNF384 | PAX6 | 9925 | Repressor | Enhancing | −0.398 | 6.90E−07 | 1.48E−04 | −0.028 | 5.94E−01 | 7.64E−03 | 2.15E−01 | 1.68E−02 | (78,79) | (92) |
| cg11871337 | chrX: 56991382 | FOXL1 | SPIN3 | 4443 | Activator | Attenuating | −0.294 | 2.18E−07 | 2.68E−04 | 0.183 | 4.32E−04 | 7.26E−01 | 9.95E−01 | 8.21E−07 | (82,83) | (93) |
| cg04665204 | chr17: 81453398 | POU3F4 | BAHCC1 | 57 922 | Dual | Invert | −0.37 | 6.80E−07 | 2.91E−04 | 0.083 | 1.12E−01 | 2.69E−05 | 1.47E−06 | 2.38E−02 | (57,61) | (52,63) |
Triplet data included the CpGs located in promoter regions (<2 kb from TSS) or distal regions (>2 kb from TSS), its location (hg38 genome), the target genes and the TFs with binding sites within 250 bp of the CpGs. Annotations included the distance between CpG and target gene (CpG–target distance), MethReg predicted roles of the TF on the target gene and the effect of DNA methylation (DNAm) on the TF. TF activity (hereafter referred to as TF) was first estimated using the VIPER software. After removing effects of CNA and tumor purity estimate in gene expression data, the robust linear model target gene expression residual ∼ DNAm group + TF + TF × DNAm group was then fitted to data to obtain estimated TF × DNAm group effect (estimate) and its P-value; the stage-wise method was used to compute the overall FDR (adj. P-value) for the term DNAm.group × TF. TF–target association in all samples was estimated using a correlation coefficient. P-values for tumor versus normal comparisons were obtained by applying Wilcoxon rank-sum tests to compare DNAm levels, TF and target gene in tumor and normal samples separately.