DANIO-CODE Gap-filling data

Zebrafish is not only a popular model for embryonic development but it is increasingly used to successfully model human disease, to screen therapeutic effects and the toxicity of drugs. Zebrafish genome is the third most complete vertebrates. Functional annotation of this genome is essential to realise the power of this model organism in diverse areas of biomedical research. DANIO-CODE international consortium aimed to annotate the functional genomic elements of the developing zebrafish embryo. We combined sets of unpublished genomics data with reanalysed published data of 37 labs and created a public resource https://danio-code.zfin.org. It holds >1150 high throughput sequencing samples genome-wide datasets, 21 assay types at 38 stages of development. This resource is freely available in a Data Coordination Centre designed for continued integration of zebrafish genomics data generated by the scientific community. The DANIO-CODE resource was used to map cis-regulatory landscape of development by integrating transcriptome, chromatin feature, DNA methylation and chromatin topology datasets. We identified over 140k cis-regulatory elements and subclassified loosely defined enhancers and promoters into novel, developmentally-relevant categories according to their activity state-associated chromatin acrhitectures and tissue- and stage-specificity. We have uncovered novel open chromatin features with striking differences in histone mark composition and target gene-associated candidate enhancer distribution between early (genome activation) active enhancers and those active during organogenesis suggesting distinct phases of developmental cis regulation. We discovered novel constitutive (COPE) and developmentally regulated (DOPE) orphan open chromatin elements lacking any histone modification marks. We have identified and characterized remarkably large, contiguous H3K27ac domains, which resemble 'super-enhancers' and appear to seed interacting intra-TAD topology domains and A-compartments during early development. Finally, we developed a novel method for comparative epigenomics, by directly comparing chromatin features such as polycomb repression-associated H3K27me3 in TADs and subTADs between zebrafish and mouse loci without direct sequence similarity and in divergent sequence lengths. We demonstrate evolutionary conservation of the precise extent of polycomb repression within gene regulatory blocks. We were able to identify zebrafish and mouse syntenic cis regulatory elements leading to cross-species functional relationship predictions without sequence similarity, thus extending the utility of zebrafish developmental genomics to that of mammalian.