A: DMRs | 5' type-3 | 3' type-2 | 3' type-3 | 5' type-1 |
 5' type-2
|
7.04
<2e-6
| 2.31 >0.12 |
11.6
<7e-11
| Â |
 5' type-3
| Â |
9.06
<4e-7
| 2.69 >0.01 | Â |
 3' type-2
| Â | Â |
6.97
<8e-6
| Â |
 3' type-1
| Â | Â | Â |
5.87
<5e-7
|
B: Genes
|
DMR
|
body
|
5’ reg
|
3’ reg
|
 TE
|
  body
|
3.28
<4e-22
| 1.12 >0.54 | 1.16 >0.44 |
  5’ reg
  +body
|
1.46
<6e-5
| 0.97 >0.87 | 1.28 >0.13 |
  5’ reg
  -body
|
0.29
<5e-15
| 0.9 >0.68 | 0.83 >0.41 |
  3’ reg
  +body
|
1.94
<6e-12
| 1.28 >0.13 | 1.02 >0.93 |
  3’ reg
  -body
|
0.33
<4e-13
| 0.94 >0.83 | 0.94 >0.83 |
- In both A and B, cells contain odds-ratios (top) and uncorrected p-values (bottom) from Fisher’s exact test, and highly significant results (p<6e-5) are in bold. A: Significance of overlaps between different types of DMR-regulatory region hits (see Figure 1d). Type-1: within 5kb of a gene but outside the body. Type-2: overlapping any of the first or last 1kb of a gene. Type-3: a type-1 or type-2 DMR for one gene that was also within the body of another gene on the opposite strand. Empty cells are either a redundant or undefined comparison, e.g. by definition 5’ and 3’ type-1 DMRs cannot overlap any part of a gene body so there were no overlaps with type-2/3. B: TEs in gene bodies and regulatory regions were predictive of DMRs in gene bodies but not regulatory regions. Significance of overlaps between genes based on whether they had a DMR in their body or regulatory region (columns) and a TE in their body, regulatory region and body (+body), or regulatory region and not body (-body, rows)
