Scripts for preprocessing both the scBS-seq data and the scRNA-seq data.

scBS-seq preprocessing

In scripts/run_bismark_only_met.sh, we get data from the adaptor-and-quality trimmed fastq files, and run bismark. We run bismark single-end, non-directional mapping (which invovles both R1 and R2). Then, we deduplicate the bismark output files, merge all 2 files, extract the methylation information with bismark_methylation_extractor and finally generate the coverage report using coverage2cytosine

variables

• input_dir=$1 # the entire path to the directory containing R1 and R2
• trim_fq1=$2 # just the file name for R1
• trim_fq2=$3 # just the file name for R2
• out_dir=$4
• sample=$5 #sample name, or data_des. Can be arbitrary. Do not use to identify a file
• recompute=$6
• multicore=$7
• genome_ref=$8 # reference genome

scRNA-seq preprocessing

In scripts/RNA_scLimeCat.sh, we preprocess the fastq files from scRNA-seq obtained in the scSTRT-seq protocol, and generate the cell-by-count matrix.

Variables:

• out_dir=$1
• R1_file_0=$2
• R2_file_0=$3
• script_dir=$4
• cell_barcode=$5 # white_list cell barcode directory
• sample_name=$6
• recompute=$7
• protocol=$8
• reference=$9 # reference genome