CAGE-seq

Warning

under construction

Integration with FANTOM5 data

From a user view point it will be more convenient when CAGE-seq data will be integrated one with other and with FANTOM5 data.

Main ideas are following:

• to build merged set of TSSs from FANTOM5 and obtained data;

• when TSSs are overlapping in FANTOM5 and obtained data, then maximally preserve FANTOM5 TSSs and in rare cases add new TSSs. This is the most sophisticated part. See its description below.

• using the joined set of TSSs to build the joined table of TSSs expressions using FANTOM5 and new data. For this purpose we are recalculating TSSs expressions using their joined set. Due to the suggested approach all FANTOM5 data will be preserved and seamlessly extended by the new data.

• annotate new TSSs using the same approach as FANTOM5. All FANTOM5 data also will be preserved and seamlessly extended by the new data.

• while FANTOM5 does not provide enhancers and their annotation for CAGE-seq data for rat, we are building a set of enhancers using both FANTOM5 and our data. to provide a unique ID for each TSS, enhancer and promoter so a user can unambiguously refer to them.

The suggested approach can be used for incremental integration of other CAGE-seq data with FANTOM5 data.

Organisms

The GTRD database and FANTOM5 contains corresponding data for different organisms.

The table below shows species in the GTRD database and FANTOM5.

Currently we include in the GTRD database CAGE-seq from FANTOM5 for 3 common species only: Homo sapiens, Mus musculus and Rattus norvegicus.

Species	GTRD	FANTOM5	Comment
Arabidopsis thaliana
Caenorhabditis elegans, nematode
Danio rerio, zebrafish
Drosophila melanogaster, fruit fly
Homo sapiens, human
Mus musculus, mouse
Rattus norvegicus, rat
Saccharomyces cerevisiae, budding yeast
Schizosaccharomyces pombe, fission yeast
Canis familiaris, dog			possibly will be added in GTRD next releases
Gallus gallus, chicken			possibly will be added in GTRD next releases
Macaca mulatta, rhesus macaque			possibly will be added in GTRD next releases

FANTOM5 data import

Table below specifies FANTOM5 files that were imported into the GTRD database.

Species	FANTOM5 directory, files	Comments
Homo sapiens	https://fantom.gsc.riken.jp/5/datafiles/reprocessed/hg38_v8/extra/ peaks - hg38_fair+new_CAGE_peaks_phase1and2.bed.gz annotation - hg38_fair+new_CAGE_peaks_phase1and2_ann.txt.gz enhancers - F5.hg38.enhancers.bed.gz expression - hg38_fair+new_CAGE_peaks_phase1and2_counts.osc.txt.gz experiment description - HumanSamples2.0.sdrf.xlsx
Mus musculus	https://fantom.gsc.riken.jp/5/datafiles/reprocessed/mm10_v8/extra/ peaks - mm10_fair+new_CAGE_peaks_phase1and2.bed.gz annotation - mm10_fair+new_CAGE_peaks_phase1and2_ann.txt.gz enhancers - F5.mm10.enhancers.bed.gz expression - mm10_fair+new_CAGE_peaks_phase1and2_counts.osc.txt.gz experiment description - MouseSamples2.3.sdrf.txt
Rattus norvegicus	https://fantom.gsc.riken.jp/5/datafiles/latest/extra/CAGE_peaks peaks - rn6.cage_peak_coord.bed.gz annotation - rn6.cage_peak_ann.txt.gz enhancers - are missing in FANTOM5 expression - rn6.cage_peak_counts.osc.txt.gz experiment description - RatSamples2.1.sdrf.txt

CAGE peaks annotation

If CAGE peak is located near with known gene TSS then it is annotated as promoter.

Also such peak is annotated with hyperlinks to corresponding transript/gene/protein identifier in the external database.

For each specie its own set of databases is used as shown in table below.

Column comments:

• Database, entity - database name and correspondin entity type (transript/gene/protein)

• FANTOM5 - indicates whether references to this database are used in annotation by FANTOM5

• GTRD, FANTOM5 - indicates whether references to this database are used in GTRD for CAGE peaks imported from FANTOM5

• GTRD, new peaks - indicates whether references to this database are used for annotation of new CAGE peaks in GTRD

Species	Database, entity	FANTOM5	GTRD, FANTOM5	GTRD, new peaks	Description, comment
Homo sapiens	Ensembl, transcript id
	Ensembl, gene id				duplicates cage_promoters2transcripts?
	Entrez, gene id
	HGNC, gene id
	FANTOM5, peak id				duplicates cage_peaks_xx.fantom5_id?
	Associated gene name Associated gene symbol Associated gene synonym Source of associated gene
Mus musculus	Ensembl, transcript id
	Ensembl, gene id				duplicates cage_promoters2transcripts?
	Entrez, gene id
	FANTOM5, peak id				duplicates cage_peaks_xx.fantom5_id?
	Associated gene name Associated gene symbol Associated gene synonym Source of associated gene
Rattus norvegicus	Ensembl, gene id
	Ensembl, transcript id				duplicates cage_promoters2transcripts?
	Ensembl_gene_name Augustus_gene_name RefSeq_transcript_id Geneid_gene_name Genscan_gene_name EST_gene_name		?

Database schema

Results of analyses of CAGE-seq data presented as a set of SQL tables.

These tables are populated for each species, where xx is suffix that specifies specie and genome build.

Table suffix	Species	Genome build
hg38	Homo sapiens, human	https://www.ncbi.nlm.nih.gov/assembly/GCF_000001405.26
mm10	Mus musculus, mouse	https://www.ncbi.nlm.nih.gov/assembly/GCF_000001635.20
rn6	Rattus norvigicus, rat	https://www.ncbi.nlm.nih.gov/assembly/GCF_000001895.5

ER diagram

Database schema for CAGE-seq data

Table cage_peaks_xx

These tables describes CAGE peaks (TSSs - Transcription Start Sites).

Field	Type	Index	Description
id	int(10)	primary	internal CAGE peak (TSS) identifier
chrom	varchar(10)		chromosome according to coresponding genome build (xx) Example: chr1.
start	int(10)		start of CAGE peak region
end	int(10)		end of CAGE peak region
strand	char(1)		TSS strand (+ or -)
abs_summit	int(10)		start of the representative TSS position
name	mediumtext -> varchar	?	name (ID) of the CAGE peak Example: hg19::chr1:564571..564600,+;hg_1.1
type	varchar(255)	?	type of CAGE peak: promoter or enhancer
organism	varchar(50)		organism. ? remove
gene_symbol	mediumtext -> varchar		gene symbol for CAGE peaks that are promoters
enhancer_id	int(10)		id of enhancer that includes this CAGE peak
fantom5_id	mediumtext -> varchar		CAGE peak id in FANTOM 5
version	int(10)		CAGE peak version
source	varchar(255)	?	Source (database or project) from which this CAGE peak was imported. Possible values: FANTOM 5

Table cage_promoters_annotation_xx

Some CAGE peaks are annotated as promoters. Corresponding links are stored in cage_promoters_annotation_xx tables.

Field	Type	Index	Description
peak_id	int(10)		internal CAGE peak (TSS) identifier
external_db	varchar(30)		External database name
external_db_id	mediumtext -> varchar		Corresponding transript/gene/protein identifier in the external database

Table cage_promoters2transcripts_xx

Some CAGE peaks are annotated as promoters. Tables cage_promoters2transcripts contain distance and id for nearest transcripts (distance <500 bp).

Field	Type	Index	Description
peak_id	int(10)	mul	internal CAGE peak (TSS) identifier
ensembl_transcript_id	varchar(30)
distance	int(11)

Table cage_enhancers_xx

Some CAGE peaks are joined into enhancers.

Field	Type	Index	Description
id	int(10)	primary	internal enhancer identifier
chrom	varchar(10)		chromosome according to coresponding genome build (xx) Example: chr1.
start	int(10)		start of enhancer region
end	int(10)		end of enhancer region
type	varchar(255)		Source (database or project) from which this enhancer was imported. Possible values: FANTOM 5 rename to sourse
name	mediumtext -> varchar		name (ID) of the CAGE peak Example: chr10:100006233-100006603

Table cage_enhancers2genes_xx

These tables describe what genes are potentially regulated by enhancers.

Field	Type	Index	Description
enhancer_id	int(10)		internal enhancer identifier
ensembl_gene_id	varchar(20)
type	enum(‘corr’,’prox’)		type of possible promoter-enhancer interaction
distance	int(11)		distance between promoter-enhancer, for type ‘prox’
corr	float		expression correlation between promoter-enhancer, for type ‘corr’
padj	float		adjusted P value for expression correlation between promoter-enhancer, for type ‘corr’