TCeA portal includes a catalog of H3K27ac ChIP-seq defined super-enhancers in 86 ENCODE cell and tissue types (Denes Hnisz et al 2013 Data S2). We studied 5 Mb core super-enhancers regions active in >20 of the 86 cell and tissue types and developed a PCA model discribing the association between nuclesome positioning and eRNA expression. The PCA model was applied onto the other tissue-specific super-enhancers regions (~372 Mb) for eRNA loci discovery. A total of ~300,000 eRNA loci were found using this methodology.

We also expanded the super-enhancer regions for eRNA loci discovery by collecting a brunch of >300 high quality H3K27ac ChIP-seq profiles from the SEdb and cistrome database. We applied the ROSE program (Whyte et al 2013) to annotate ~320Mb irredundant super-enhancer regions outside those in the 86 samples. TCeA identified ~200,000 eRNA loci in these regions. Besides these, TCeA also included ~63,000 eRNAs from typical enhancers annotated by the FANTOM project. This is an updated dataset of Chen et al 2018.

All the sample, enhancer, and eRNA annotations can be downloaded below:

Download the 377 Mb super-enhancer regions in 86 human cell and tissue types (From Denes Hnisz et al 2013 Data S2)
Download the 5Mb core super-enhancer regions in > 20 cell and tissue types
Download the ~300,000 eRNA loci identified in the ENCODE dataset
Download H3K27ac ChIP-seq sample list from SEdb/cistrome database
Download the 320 Mb putative super-enhancer regions discivered in SEdb/cistrome database
Download the ~200,000 eRNA loci identified in the SEdb/cistrome dataset
Download the ~63,000 eRNAs loci from typical enhancers

TCeA includes ~9,300 TCGA tumor samples and ~700 matched normal samples. The TCGA patient IDs are used as tickers in TCeA. Check here for details of TCGA sample barcodes. For example, the TCeA ticker "TCGA-BL-A13J_tumor" and "TCGA-BL-A13J_normal" represent the tumor and normal samples from the BLCA patient labeled with TCGA patient ID "TCGA-BL-A13J". We used only one tumor (or normal) sample for patients with multiple runs. We choose the primary tumor samples over the metastatic ones and choose the one with the smallest portion ID when multiple portions of the same sample are available in TCGA RNA-seq dataset. All samples used in TCeA are considered as in the whitelist by TCGA consortium. The details can be downloaded below:

A total of 934 paired-fastq files of Cancer Cell Line Encyclopedia (CCLE) project were downloaded from the legacy UCSC Cancer Genome Hub (CGhub), which can now be found on EBI. The CGhub IDs (e.g. G20490.HCT_116.2 for the HCT116 cell line) are used as tickers in TCeA data portal. The cancer type of each cancer cell line (e.g. COAD for the HCT116 cell line) is matched to those used in TCGA or ICGC projects. Nine of the 934 fastq files were found to be truncated or with low mappability after multiple download attempts and thus dropped from TCeA dataset. Details of the 925 cancer cell lines included in TCeA can be downloaded below:

Download CCLE cell line annotation

TCeA combines the GTEx "submitted_subject_id" and "histological_type" as tickers. For example, the sample "GTEX-T6MO__Ovary" represents the ovarian sample from the individual with GTEx id "GTEX-T6MO". Notably, samples with multiple records (or runs) are pooled in the TCeA portal. Details of the GTEx samples included in TCeA can be downloaded below: