Table 2.
Validation of method-specific DMRs for MeDIP, MethylCap and RRBS
| DMR location | Description | Experimental validation | MeDIP | MethylCap | RRBS |
|---|---|---|---|---|---|
| MeDIP-specific DMR chr10:88,149,016-88,149,732 | Intergenic CpG island ~30kb up-stream of GRID1, partial overlap with degenerate L1 element | HUES6: 38/56 (68%) methylated CpGs HUES8: 26/44 (59%) methylated CpGs ➔ insignificant (p=0.41) | hypermethylated (q=1.1E-04) | insignificant (q=0.59) | insignificant (q=0.43) |
| MeDIP-specific DMR chr16:31,142,904-31,143,799 | CpG island overlapping with the terminal exon of TRIM72 | HUES6: 342/362 (94%) methylated CpGs HUES8: 466/523 (89%) methylated CpGs ➔ marginally hypermeth. (p=0.0051) | hypermethylated (q= 1.2E-05) | insignificant (q=0.73) | insufficient coverage |
| MeDIP-specific DMR chr1:211,290,079-211,290,896 | CpG island overlapping with the putative promoter region of RPS6KC1 | HUES6: 53/60 (88%) methylated CpGs HUES8: 45/50 (90%) methylated CpGs ➔ insignificant (p=1.0) | hypermethylated (q= 3.0E-06) | insignificant (q=0.97) | insignificant (q=0.29) |
| MethylCap-specific DMR chr20:29,526,646-29,527,380 | CpG island overlapping with the putative promoter region of REM1 | HUES6: 5/72 (7%) methylated CpGs HUES8: 78/84 (93%) methylated CpGs ➔ hypomethylated (p= 1.4E-30) | insufficient coverage | hypomethylated (q= 1.8E-09) | insufficient coverage |
| MethylCap-specific DMR chr2:151,825,938-151,826,902 | CpG island overlapping with the putative promoter region of RBM43 and a known copy-number variation | HUES6: 161/208 (77%) methylated CpGs HUES8: 9/104 (9%) methylated CpGs ➔ hypermethylated (p= 3.3E-33) | insignificant (q=0.18) | hypermethylated (q= 7.3E-09) | insufficient coverage |
| MethylCap-specific DMR chr13:44,348,934-44,349,700 | Intergenic CpG island ~60kb up-stream of NUFIP1, partial overlap with degenerate Alu element | HUES6: 80/88 (91%) methylated CpGs HUES8: 41/79 (52%) methylated CpGs ➔ hypermethylated (p= 1.2E-08) | insignificant (q=0.40) | hypermethylated (q= 8.3-07) | insufficient coverage |
| RRBS-specific DMR chr3:186,889,821-186,890,200 | CpG island overlapping with an internal exon and intron of IGF2BP2 | HUES6: 5/90 (6%) methylated CpGs HUES8: 88/90 (98%) methylated CpGs ➔ hypomethylated (p= 4.3E-42) | insufficient coverage | insignificant (q=0.18) | hypomethylated (q= 3.5E-40) |
| RRBS-specific DMR chr3:32,609,320-32,609,612 | Intergenic CpG island ~20kb up-stream of DYNC1LI1 | HUES6: 41/121 (34%) methylated CpGs HUES8: 130/143 (91%) methylated CpGs ➔ hypomethylated (p= 3.5E-23) | insufficient coverage | insignificant (q=0.52) | hypomethylated (q= 2.9E-26) |
Eight method-specific DMRs were selected based on the DNA methylation maps of MeDIP, MethylCap and RRBS and experimentally validated using bisulfite sequencing. The validation candidates were manually chosen from the list of CpG islands (Figure 5), such that each region is identified as DMR by only one method, while the other two methods do not detect any significant (or suggestive) difference in DNA methylation between the two human ES cell lines (HUES6 and HUES8). The validation was performed by clonal bisulfite sequencing with an average of 11 clones per region in each of the two human ES cell lines. The p-values in column 3 were calculated from the clonal bisulfite sequencing data using Fisher’s exact test, based on the DNA methylation levels of individual CpGs. The q-values in column 4 to 6 were derived from the DNA methylation maps as described in the Methods section. One out of three MeDIP-specific DMRs, three out of three MethylCap-specific DMRs and two out of two RRBS-specific DMRs were confirmed by clonal bisulfite sequencing data (bold print). All genomic coordinates refer to the amplicon on which the validation was performed, which was chosen such that it overlaps with the most differentially methylated region within the selected CpG islands. The coordinates are given relative to the NCBI36 (hg18) assembly of the human genome. A detailed documentation of the validation experiments is available online (Supplementary File 1).