Promoter_similarities

Preliminary datasets

EPD_promoter_analysis.R

FANTOM3_4_datasets.R

Correlation of activity between proximal promoters

1 - FANTOM5_prom_ID.R

Identify human "True TSS clusters" from FANTOM5 dataset (Data/TSS_human.bed) as the onea with a minimum score value and associated with an ENSEMBL element (Data/ensembl_genes_grch37.txt: downloaded from BioMart)
Outputs are in Tables/h37_tss_selected_gr.bed and in Tables/clusters_selected_genes.csv

2 - fantom5_links.R

Get the links where the FANATOM5 CAGE files are, using CAGEr package.
The links are in Tables/FANTOM5_files.txt and the whole sample information table of the in Tables/FANTOM5_files_info.tsv

3 - FANTOM5_files.sh

Bash script to download the files from the links in Tables/FANTOM5_files.txt in parallel, saving the in folder Data/FANTOM5_files/

4 - init_CAGEfiles.R

Loading the CAGE files (in parallel), merging data from the same cell_type and overlapping with the selected TrueTSS regions (Tables/h37_tss_selected_gr.bed).
Output to Tables/true_tss_activ_tissue.rds, with a dataframe for each cell type/tissue in each element from the list.

5 - tissue_activity.R

With Tables/true_tss_activ_tissue.rds as input, firstly it restricts to the promoters in Tables/clusters_selected_genes.csv.
Secondly, summarizes data from all samples, to filter the active promoters along multiple samples (Tables/active_promoters.tsv) and those highly tissue specific (Tables/tissue_restricted/prom.tsv).

6 -distances.R

Calculate distance to the nearest active promoter in each gene for each promoter in Tables/active_promoters.tsv.
Output in Tables/active_promoters_df.csv

7 -closed_prom.R

Firstly, identify the IC95 of the width of the selected TSS clusters.
Select the pairs of promoters from each gene that are within this genomic distance.
proximal_promoters.rds contains the output.

8 -closed_prom_corr.R

Variance of relative activity of promoters in proximal promoters (Tables/proximal_promoters.rds ) along the different human cell types (Tables/true_tss_activ_tissue.rds).
Control is assessed by permuting the pairs of promoters