This tool is intended to perform demultiplexing when no index reads are available. It works by determining the minimal required barcode length, searching for the adapter sequence in the read and, if successful, attempting to match the subsequent barcode with one of the sample barcodes. The output is written to FASTQ files.
The adapter and the sample names and barcodes are provided in a YAML file (samples.yaml). This file needs to be created for each demultiplexing run. Additional configuration for the adapter matching and barcode length settings is located in the file demux_recovery.yaml.
The script requires Biopython to run, and was written with Python 3.6.
./demux_recovery.py --input_fastq Undetermined_S0_L001_R1_001.fastq.gz --samples_file samples.yaml --output_dir . --log_level DEBUGSingle-threaded usage can take a very long time if there is a large number of reads. The tool can be run in parallel using the includes Snakemake pipeline. Copy the Snakefile, config.yaml and samples.yaml to the location of the FASTQ file, adjust the config.yaml and run snakemake.
Cluster submission scripts for SGE are included, but will most likely need to be adapted.