snATAC_preprocess:

A tool for demultiplex snATAC into samples and id nuclei by combining barcodes.

Usage

snATAC_preprocess.py --help

Usage: snATAC_preprocess.py [-h] [-a I1.fastq] [-b I2.fastq] [-c R1.fastq] [-d R2.fastq]

Decomplex single-cell ATAC-seq barcode allowing mismatch.

Options:
  -h, --help            show this help message and exit
  -a I1, --i1=I1        I1.fastq.gz
  -b I2, --i2=I2        I2.fastq.gz
  -c R1, --r1=R1        R1.fastq.gz
  -d R2, --r2=R2        R2.fastq.gz
  -m S1, --s1=S1        Sample 1 using barcode set1
  -n S2, --s2=S2        Sample 2 using barcode set2
  -x XMISMATCH, --xmismatch=XMISMATCH
                        max allowed mismatch (default 2).
  -l BARCODES, --barcode=BARCODES
                        a single barcode library text file.
  -o OUTDIR, --outdir=OUTDIR
                        output dir.
  --version=VERSION     

Test

 ./snATAC_preprocess.py  -a ./test/I1.fastq.gz -b ./test/I2.fastq.gz -c ./test/R1.fastq.gz -d ./test/R2.fastq.gz -m A -n B -l ./barcodes/barcodes.txt -o ./test/

Two rounds of multiplexing in snATAC-seq

Libraries are sequenced paired end. Each sequencing read is assigned to 4 barcodes on 2 Index reads. The index reads are 43 (Index 1) and 37 (Index 2) reads long.

Two rounds:

1. During tagmentation (Tn5 treatment) p/r barcodes are integrated
2. During PCR, i barcodes are integrated

PCR after FACS Positions for the barcodes are: p/r7 i7 i5 p/r5.

• p/r7 1-8 (Index read 1 or I1)
• i7 36-43 (Index read 1 or I1)
• i5 1-8 (Index read 2 or I2)
• p/r5 30-37 (Index read I2)

Two sample per seq run

In each seq run, two samples are mixed and they are multiplexed by two sets of barcodes. Usually, we use the same Tn5 but different PCR barcodes for Sample A and B. So:

• Same: p/r7 and p/r5
• Diff: i7 (for R1 reads) and i5 (for R2 reads)

So in snATAC preprocessing, the total fastqs will have two levels demultiplexing:

1. demultiplex to Sample A and Sample B
2. demultiplex to individual cell/nucleus

By default, our script will demultiplex each seq run to level 2 (single nuleus level for each sample).