Table 1. Genomic annotation-centred analysis of differential methylation after early gestational famine exposure.
Genomic annotations* | Covered with RRBS/total in genome (%) | Pnominal | PFDR |
---|---|---|---|
Non-CGI, ‘bona fide’ promoters† | 2,024/7,014 (28.9%) | 9.1 × 10−4 | 0.026 |
Enhancers‡ | 6,207/59,466 (10.4%) | 1.9 × 10−3 | 0.026 |
DNaseI/FAIRE-seq regions§ | 79,728/590,252 (13.5%) | 4.4 × 10−3 | 0.036 |
Middle exons | 1,570/17,848 (8.8%) | 5.8 × 10−3 | 0.036 |
Developmental enhancers type I|| | 922/5,118 (18.0%) | 6.5 × 10−3 | 0.036 |
‘bona fide’ CGI shores¶ | 27,688/88,871 (31.2%) | 0.012 | 0.053 |
Non-coding RNA# | 59/718 (8.2%) | 0.015 | 0.053 |
Conserved regions** | 1,386/165,937 (0.8%) | 0.016 | 0.053 |
CGI shores | 67,811/319,509 (21.2%) | 0.017 | 0.053 |
3′UTR | 2,909/21,004 (13.8%) | 0.035 | 0.085 |
Non genic CGI†† | 41,023/129,049 (31.8%) | 0.036 | 0.085 |
‘Bonafide’ CGI border | 22,777/88,074 (25.9%) | 0.036 | 0.085 |
Developmental enhancer type II | 320/2,287 (14.0%) | 0.078 | 0.15 |
CGI | 113,673/343,925 (33.1%) | 0.078 | 0.15 |
Introns | 61,816/201,640 (30.7%) | 0.080 | 0.15 |
hESC bivalent chromatin domains | 1,741/1,797 (96.9%) | 0.16 | 0.28 |
Bonafide CGI | 35,271/44,439 (79.4%) | 0.20 | 0.32 |
Cell-type specific gene promoters | 2,106/2,372 (88.8%) | 0.21 | 0.32 |
First exons | 13,507/51,497 (26.2%) | 0.25 | 0.36 |
Promoters | 16,904/23,689 (71.4%) | 0.26 | 0.36 |
HSC bivalent chromatin domains | 2,779/2,910 (95.5%) | 0.28 | 0.36 |
Imprinted promoters | 42/46 (91.3%) | 0.29 | 0.36 |
‘Bona fide’ CGI promoter | 14,880/16,674 (89.2%) | 0.32 | 0.37 |
CTCF insulators from CD4+ cells | 4,396/28,661 (15.3%) | 0.32 | 0.37 |
Imprinted DMRs | 6/14 (42.9%) | 0.33 | 0.37 |
Putative metastable epialles | 29/38 (76.3%) | 0.43 | 0.47 |
Variably methylated regions | 56/227 (24.7%) | 0.55 | 0.57 |
Promoters cancer genes | 795/888 (89.5%) | 0.63 | 0.63 |
DMR, differentially methylated region; hESC, human embryonic stem cell; RRBS, reduced representation bisulfite sequencing; UTR, untranslated region.
The coverage of the genomic annotations is in line with the enrichment of RRBS for GC-rich regions, but also indicates a good coverage across annotations, including those relatively GC poor. P values were obtained with GlobalTest.
*More details on the genomic annotations can be found in the Methods.
†Promoters without CGIs but with a relatively open chromatin state22.
‡Enhancers characterized by H3K4me1, non-overlapping with promoters23.
§Regions with an open chromatin state as defined by DNaseI and FAIRE-seq signals (UCSC track ENCODE).
||Enhancers active during first stages of blastocyst development24.
¶Shores of the so-called bona fide CGI, CGI island with an ubiquitously open chromatin structure; oe>0.6, GC%>50% and length >700 bp22.
#Body of various type of non-coding RNAs.
**Conserved regions outside promoters, CGIs, exons and UTR.
††CGI>10 kb from gene.