Atlas of regulatory elements

Developmental cis-regulatory elements of zebrafish, identified by the DANIO-CODE consortium from ATAC-seq and histone-modification ChIP-seq across embryonic development, for the danRer11 (GRCz11) and danRer10 (GRCz10) assemblies.

View danRer11 in the UCSC Genome Browser View danRer10 in the UCSC Genome Browser

If you use these annotations, please cite:

Baranasic D, Hörtenhuber M, Balwierz PJ, et al. Multiomic atlas with functional stratification and developmental dynamics of zebrafish cis-regulatory elements. Nature Genetics 54, 1037–1050 (2022). https://doi.org/10.1038/s41588-022-01089-w

Element classes

PADREs: Predicted ATAC-supported developmental regulatory elements: ATAC-seq peaks reproducible between replicates. At the five stages with histone-modification data, each PADRE is labelled with its chromatin state at that stage.
cPADREs: Consensus PADREs: the strict consensus set of regions open in at least two neighbouring developmental stages. The five stage files hold the same regions, each labelled with the chromatin state at that locus in that stage; a label alone does not mean that the region is open at that stage, which is what the stage's PADREs show.
DOPEs: Dynamic orphan predicted elements: cPADREs that are quiescent (no promoter, enhancer or Polycomb marks) at all five stages. Each stage file lists the DOPEs open at that stage, that is, overlapping a PADRE of that stage; the pooled file lists all of them. The definition does not exclude elements that are open at every stage, so the DOPEs include the COPEs.
COPEs: Constitutive orphan predicted elements: distal PADREs open at every stage and quiescent at every stage.
Chromatin states: Genome-wide segmentation into ten states with ChromHMM, from H3K4me3, H3K4me1, H3K27ac and H3K27me3, in 200-bp bins, at each of the five stages.
Prim-5 cell types: Prim-5 PADREs assigned to cell types from single-cell ATAC-seq. The state numbers are those of McGarvey et al., Cell Genomics 2, 100083 (2022). A region can be specific to several cell types, so the combined file has more rows than distinct regions.
SOM classes danRer10: Distal cPADREs (more than 500 bp from a CAGE promoter) clustered by their accessibility over development into a 4 × 4 self-organising map, giving 16 classes.
H3K27ac ensembles danRer10: Large clusters of H3K27ac at the Dome stage, called with ROSE. They resemble super-enhancers but appear earlier and are more numerous.
eRNA-defined enhancers danRer10: Intergenic and intronic regions with bidirectional enhancer RNA in nuclear CAGE, called with CAGEfightR.
Transgenic reporter-validated enhancers: A manually curated catalogue of published enhancers tested in zebrafish transgenic reporter assays.
Mouse H3K27me3 projection danRer10: Mouse E10.5 H3K27me3 projected onto zebrafish through multispecies alignment anchors, and the zebrafish 1-kb bins projected onto mouse mm10.

Chromatin states

PADREs, cPADREs and the segmentations are labelled with one of ten states and coloured accordingly in the files and in the genome browser.

	State	Meaning
	`1_TssA1`	Active TSS 1
	`2_TssA2`	Active TSS 2
	`3_TssFlank1`	TSS flanking region 1
	`4_TssFlank2`	TSS flanking region 2
	`5_EnhA1`	Active enhancer
	`6_EnhFlank`	Enhancer flanking region
	`7_EnhWk1`	Primed enhancer
	`8_Pois`	Poised element
	`9_ReprPC`	Polycomb-repressed region
	`10_Quies`	Quiescent

The same analysis re-run natively on the current assembly.

By stage

Stage	Chromatin states	PADREs	cPADREs	DOPEs
Dome 4.3 hpf	239,617 BED · bigBed	85,764 BED · bigBed	133,147 BED · bigBed	4,897 BED · bigBed
75%-epiboly 8 hpf	296,414 BED · bigBed	93,408 BED · bigBed	133,147 BED · bigBed	5,864 BED · bigBed
5-9 somites 12 hpf	340,022 BED · bigBed	123,340 BED · bigBed	133,147 BED · bigBed	7,388 BED · bigBed
Prim-5 24 hpf	300,667 BED · bigBed	122,476 BED · bigBed	133,147 BED · bigBed	7,704 BED · bigBed
Long-pec 48 hpf	1,121,204 BED · bigBed	148,887 BED · bigBed	133,147 BED · bigBed	6,471 BED · bigBed

Across stages

Set	Regions	Download
COPEs	2,202	BED · bigBed
DOPEs, all stages pooled	10,726	BED · bigBed
Transgenic reporter-validated enhancers (Supplementary Table 10)	246	BED · bigBed

Prim-5 cell types

Cell type	State	Regions	Download
All cell types		54,230	BED · bigBed
Epidermis	1	3,273	BED · bigBed
Muscle II	3	5,278	BED · bigBed
Optic vesicle	4	2,277	BED · bigBed
CNS	5	2,449	BED · bigBed
Muscle I	10	1,894	BED · bigBed
Epidermis / olfactory / pronephric	12	3,573	BED · bigBed
Muscle / tailbud	13	3,486	BED · bigBed
Spinal cord	14	1,831	BED · bigBed
Epidermis II	16	7,085	BED · bigBed
CNS II	18	5,127	BED · bigBed
Neural crest	20	2,532	BED · bigBed
State 21 (labelled endothelium in danRer11)	21	2,258	BED · bigBed
Blood	23	1,357	BED · bigBed
Midbrain	24	3,146	BED · bigBed
Pharyngeal mesoderm / tailbud	25	3,083	BED · bigBed
Tailbud / spinal cord	26	1,512	BED · bigBed
Differentiating neurons	29	4,069	BED · bigBed

The version analysed in the paper; the numbers in the paper refer to these files.

By stage

Stage	Chromatin states	PADREs	cPADREs	DOPEs
Dome 4.3 hpf	183,504 BED · bigBed	87,061 BED · bigBed	141,752 BED · bigBed	4,658 BED · bigBed
75%-epiboly 8 hpf	263,984 BED · bigBed	106,346 BED · bigBed	141,752 BED · bigBed	6,750 BED · bigBed
5-9 somites 12 hpf	323,970 BED · bigBed	132,605 BED · bigBed	141,752 BED · bigBed	7,867 BED · bigBed
Prim-5 24 hpf	321,377 BED · bigBed	125,145 BED · bigBed	141,752 BED · bigBed	7,499 BED · bigBed
Long-pec 48 hpf	1,127,452 BED · bigBed	156,950 BED · bigBed	141,752 BED · bigBed	6,269 BED · bigBed

Across stages

Set	Regions	Download
COPEs	2,109	BED · bigBed
DOPEs, all stages pooled	11,044	BED · bigBed
Transgenic reporter-validated enhancers	248	BED · bigBed

Prim-5 cell types

Cell type	State	Regions	Download
All cell types		58,408	BED · bigBed
Epidermis	1	3,246	BED · bigBed
Muscle II	3	5,377	BED · bigBed
Optic vesicle	4	2,305	BED · bigBed
CNS	5	2,445	BED · bigBed
Muscle I	10	1,888	BED · bigBed
Endothelium	11	5,906	BED · bigBed
Epidermis / olfactory / pronephric	12	3,674	BED · bigBed
Muscle / tailbud	13	3,498	BED · bigBed
Spinal cord	14	1,818	BED · bigBed
Epidermis II	16	7,249	BED · bigBed
CNS II	18	5,172	BED · bigBed
Neural crest	20	2,572	BED · bigBed
Blood	23	1,423	BED · bigBed
Midbrain	24	3,171	BED · bigBed
Pharyngeal mesoderm / tailbud	25	3,075	BED · bigBed
Tailbud / spinal cord	26	1,515	BED · bigBed
Differentiating neurons	29	4,074	BED · bigBed

Other analyses

Set	Regions	Download
H3K27ac ensembles	1,649	BED · bigBed
eRNA-defined enhancers (nuclear CAGE)	10,749	BED · bigBed
Mouse E10.5 H3K27me3 projected onto zebrafish (signal)		bigWig
Zebrafish 1-kb bins projected onto mouse mm10	129,448	BED · bigBed

SOM classes of distal cPADREs

Class	Regions	Described in the paper	Download
1	5,004		BED · bigBed
2	8,735		BED · bigBed
3	6,605		BED · bigBed
4	8,714	active early, then decommissioned	BED · bigBed
5	10,097		BED · bigBed
6	11,360	active from zygotic genome activation onwards	BED · bigBed
7	6,176		BED · bigBed
8	5,915		BED · bigBed
9	10,789		BED · bigBed
10	8,449		BED · bigBed
11	4,523		BED · bigBed
12	3,448		BED · bigBed
13	9,999		BED · bigBed
14	14,902	late elements	BED · bigBed
15	3,999		BED · bigBed
16	4,371		BED · bigBed

Supplementary tables

Tables of the paper as tab-separated text, and as BED where they hold genomic coordinates. The complete workbook is available from the article page.

Table	Assembly	Rows	Download
Table 2. Number of transcripts in the DANIO-CODE transcript sets		3	TSV
Table 3. Novel lncRNA and TUCP genes (the 382 in the published table)	danRer10	382	BED
Table 5. CRISPR guide RNAs and ddCt values of promoter validation	danRer10	114	TSV
Alternative promoters of multi-promoter genes (the intended Table 6)		1,431	TSV
Table 7. Mouse multi-promoter genes	mm9	2,250	TSV
Table 8. KEGG pathways enriched in mouse multi-promoter genes		62	TSV
Table 9. GO enrichment of Tead3a/b motif target promoters		178	TSV
Table 10. Transgenic reporter-validated enhancers, with PubMed ID and expression pattern	danRer11	246	BED
Table 11. Prim-5 transgenic enhancers compared with cell types and chromatin states	danRer10	155	TSV · BED
Table 12. Genes associated with COPEs and DOPEs		60	TSV
Table 14. Shared motifs of mouse DNase and zebrafish ATAC peaks within and across TADs (KS tests)		3	TSV
Table 15. Genes associated with the most conserved H3K27ac ensembles	danRer10	71	TSV

File format

BED9 (chromatin states, PADREs, cPADREs, cell types): chromosome, start, end, name (chromatin state, or cell type and state number), score, strand (always .), thickStart, thickEnd, itemRgb (the colour of the state).
BED6 (DOPEs, COPEs, SOM classes, enhancer sets): chromosome, start, end, name, score and strand.
BED4 (1-kb bins projected onto mouse): chromosome, start, end, and the mm10 region with its projection score as the name.
Coordinates are 0-based and half-open, as in all BED files. The annotation BED files are coordinate-sorted.
The supplementary-table BED files are BED6 followed by the table's remaining columns, in the row order of the published table; the TSV files keep the table's header.
Chromosomes use UCSC names (chr1, chrUn_KN149861v1). Ensembl names the same sequences 1, KN149861.1, …
danRer11: 256 regions that ran up to 70 bp past the end of a chromosome or scaffold were clipped to it, so that every file loads in genome browsers and interval tools.
danRer10: the BED files were converted from the published bigBed files with UCSC bigBedToBed.
In the danRer10 eRNA-defined enhancers, 1,993 regions are listed twice.