Single cell ATAC coverages - settylab/single-cell-primers GitHub Wiki

Coverage plots

Coverage plots are a useful visualization to understand the accessibility changes at different loci. Coverage plots for a subset of clusters from the PBMC data at the RUNX1 locus is shown below.

drawing

Generating Coverage plots

The notebook here details the generation of coverage plots from single-cell ATAC data. A bigwig file is generated for each cluster / cell-type which can then be browsed either using IGV or [pyBigwig][https://github.com/deeptools/pyBigWig].

The following are the parameters necessary for using the notebook

# RNA and ATAC anndata objects
atac_ad

# Output and scratch directories
out_dir
tmp_dir 


# Grouping parameters 
# Sample field in atac_ad.obs (Useful in case your anndata has multiple samples)
sample_key
# Cell type field in atac_ad.obs (A bigwig will be generated for each group of cells)
celltype_key


Celltype subset - Specify if you only want coverages for a subset of cell types
celltype_subset = None

# Fragment files
fragment_files = dict()
# Add an entry for each sample 


# Chromsizes [Path to chromosome sizes file downloaded from UCSC Genome Browser]
# For human:
chromsizes = '/fh/fast/setty_m/grp/lab-datasets/bonemarrow-tcell-dep-multiome/cr-arc-results/hg38.chrom.sizes'

For each cell type, the following bigwig files will be generated

  1. <celltype>_All.bw: Coverage using all fragments
  2. <celltype>_NFR.bw: Coverage with NFR fragments (fragment length < 147)
  3. <celltype>_NUC.bw: Coverage with nucleosomal fragments (fragment length > 147)

Note

Make sure all the necessary python packages and dependencies are installed as detailed in https://github.com/settylab/single-cell-primers/wiki#environment-set-up

⚠️ **GitHub.com Fallback** ⚠️