Skip to content

Demultiplexing for Fiberseq smk

Jump to bottom
StephanieBohaczuk edited this page Aug 29, 2022 · 1 revision

If you are running the Fiberseq-smk pipeline on a multiplex sample, a demultiplexed CCS file containing reads from a single barcode should be provided. For example, if you multiplexed four samples, you will run four instances of fiberseq-smk, one for each barcode. Here's how to generate those CCS files.

  1. Unzip compressed files to extract your subreads and .pbi file for your subreads: tar -zxvf <your zipped .tar.gz file> *subreads.bam*

  2. Run ccs (see ccs.how for more details, may want to chunk file for faster processing). Make sure to include -hifi kinetics: ccs <subread file> <output file name> --hifi-kinetics

  3. Combine chunks if you used them, then run lima to demultiplex your samples. See lima.how for more info:

lima <ccs file> \ 
<fasta file containing barcode sequences> \
--hifi-preset SYMMETRIC \ # uses hifi settings for determining barcode identity, requires for the ccs read to be barcoded on both ends
--split-bam \ #splits the output bam file by barcode
--split-named \ #uses the barcode names from the barcode.fa file
--split-subdirs \  #splits the output bams into different subdirectories. This part can be omitted if you prefer to have everything in a single directory.
  1. Files are ready for fiberseq-smk pipeline See fibersq-smk github for more details. https://github.com/StergachisLab/fiberseq-smk

Example config file for demultiplexed sample:

PS00118: /mmfs1/gscratch/stergachislab/bohaczuk/data/CATCH_processing/21.12.16/for_fiberseq-smk/m64076_211222_124721.subreads.bam 
ref: /mmfs1/gscratch/stergachislab/bohaczuk/data/target_practice/hg38.REs.fa 
ccs: /mmfs1/gscratch/stergachislab/bohaczuk/data/CATCH_processing/21.12.16/for_fiberseq-smk/m64076_211222_124721.demux.bc1003--bc1003.bam
# ccs: field should be your DEMULTIPLEXED ccs bam file, the output from lima. Run each demultiplexed ccs bam in a different smk instance, 
# replacing this file accordingly. The subread file will be identical for all
env: "fiberseq-smk"

Run fiberseq-smk with your config file

snakemake --profile profile/checkpoint --configfile config/PS00118.yml -p

Clone this wiki locally