Enhancer-Promoter Linkage Atlas for the Mouse Adipose tissue generated with Enhlink. Insights into Type II Diabetes Mechanisms and Covariate-Specific Regulation

These atlases were generated using single-nucleus ATAC-seq datasets from mouse Islet and Adipose tissues. The aim of the single-cell study on the Islet and Adipose tissues was to understand the genetic-specific mechanisms involved in the development of type II diabetes. The researchers created cell-type specific enhancer-promoter atlases for both tissues, identifying which linkages were specific to the strain, diet, and sex and which intersected with the pcHI-C datasets generated from the same tissues. <br> <strong>File descriptions</strong> Each BEDPE file contains genomic links using the 6 first columns to define the enhancer (three first columns) and the promoter (the following three columns). Also, each file refers to a specific cell type written in the file name. The extension pchic refers to the links intersecting the ilset pcHI-C data. The extension unique refers to link unique within a cell type and not found in any of the other cell types.the extension 2ndOrder refers to covariate-specific links with the covariate ID described in one of the columns. Finally, the batch.tsv file describes the covariates associated with each promoter activity. <strong>Methods</strong> We independently processed each cell type of the tissue, considering library ID, sex, diet, and strain as covariates. We applied a p-value cutoff of 0.01 with the secondOrder option and used only 66% of the features, a maximum of 4 features, a depth of 2, and a downsampling size of 15,000 cells to obtain covariate-specific linkages for each tree. We generated a binary cell x promoter sparse matrix by setting 1 for a given gene g and cell c if at least one read from c fell into the promoter regions of g, and used the ATACMatUtils command with the -use_symbol option from the ATACdemultiplex package to create the matrix from the BED file containing the reads and barcode IDs. Finally, we used the enhtools software with the -intersect3 option to intersect the obtained linkage of each cell type with the pcHI-C links of either the Adipose or the Islet tissues. <br> https://gitlab.com/Grouumf/enhlinktools