This script is used for demultiplexing paired end FASTQ files based on a list of known barcodes and known barcode coordinates. This was designed to work with the CEL-seq2 single cell protocol read structure.

This script may not work on macOS due to a limit in how many files can be open. If you encounter this problem try using iTerm2 or increasing the opened files limit.

To install these scripts, clone this repository and link the scripts to a location in the system PATH.

git clone https://github.com/Shians/paired_fastq_split.git
cd paired_fastq_split
ln -s paired_fastq_split.py /usr/local/bin/ # Change /usr/local/bin as required

Do not delete or move scripts after linking.

Help documentation can be found using

paired_fastq_split.py -h

which returns the call signature

usage: paired_fastq_split.py [-h] -file1 FILE1 -file2 FILE2 -barcodes BARCODES
                      [-bc_pos BC_POS BC_POS] -prefix PREFIX [-r2bc]

A basic call looks like

paired_fastq_split.py -file1 exp_R1.fastq.gz -file2 exp_R2.fastq.gz -barcodes barcodes.txt -prefix exp

by default we assume that the barcodes are on read 1 in positions 7 to 16 (length 8). The output will be a pair of files for each barcode with the filename {prefix}_{barcode}_R(1/2).fastq along with a pair of files for reads that did not have an exact match to any barcode.

The barcode file should be plaintext with a unique barcode on each line

GTAGCTCA
TAGACCAC
GGTCTATG
GTCCGAAT
...