TiED (https://lcbb.swjtu.edu.cn/TiED), a free, web-accessible database, provides information about enhancers, eRNAs and interaction maps of enhancer, TF and gene according to High-throughput sequencing data. It currently covers 53924 enhancers (40101 TS enhancers, 1241 UE enhancers), 126 transcription factors (32TS TF), 9145 target genes of enhancers (3871 TS gene) across 10 tissues (Adrenal, Brain, Breast, Heart, Liver, Lung, Ovary, Placenta, Skeletal Muscle, Kidney). Furthermore, experimentally validated expression data of TF and gene is also collected. Users can use the “Search by region” function to retrieve the enhancer within the “region range” provided. It is also accessible to retrieve the TF, tissue or gene interacting with enhancer via “search by name” function. In addition, there is also a “TS browse” button at the top of the web page, offering approach to explore TS TFs or TS genes in each tissue by clicking 10 different tissues.
To screen the tissue-specific enhancer, the following work was done. Enhancers was identified based on four recognized features: H3K27ac, H3K4me1, H3K4me3 and DHS. Their Chip-seq data of 10 tissues was downloaded from NIH roadmap and ENCODE. Histone-binding regions were obtained by MACS (p<0.0001). Signals of histone modification were normalized using bwtool and deeptools. Regions, with two characters including higher H3K4me1 mean normalized signal than H3K4me3 and within 2kb far away from DHS, H3K27ac and H3K4me1 peaks, were selected as candidate enhancer. And they were compared with the reference genome (GENCODE.v19). Enhancers that overlapped with upstream 1 kb from the TSS of the protein coding gene were removed and the rest were our ultimate extra-gene enhancers. eRNA was defined as CAGE Tag Clusers (TCs) transcribed from enhancers. Tissue specificity of enhancer is calculated in accordance with Entropy formula.
Within the 100kb away from the enhancer, the protein coding gene closest to the enhancer is defined as its target gene. The expression (RPKM) and specificity value(DPM) of genes were obtained from GTEx and Pagenbase respectively. We downloaded the highly conserved TFBS based on ChIP-seq data of ENCODE project from UCSC database. And it includes all the TFBS data extracted from the “Txn Factor Chip”, the track which combines TFBS from many various cell line via the UCSC Table Browser. Regions 2kb upstream and 2kb downstream of each enhancer were identified as the putative transcription factor binding regions for each enhancer.