/viral-asapseq-snakemake

Primary LanguagePythonMIT LicenseMIT

viral-asapseq-snakemake

Upstream processing for single cell ASAPseq for viral vector detection. This pipeline is based on a previous pipeline that was designed for processing of sequencing techniques with antibody derived tags (see scc-proc)

Necessary files

1) config.yaml file

Example configuration is shown below.

  • If ATAC mode is enabled, the sample name MUST match with the sample name used to the demultiplex the fastq files during cellranger-atac mkfastq.
  • Multiple samples can be added to the config assuming they follow the same antibody mixture and bead setup.
  • All fastq files must be in the fastq_dir path
    • Specific filenames are provided in the samples attribute of the yaml file that correspond to the ADT files
    • Order of fastq files in the samples attribute corresponds to the cbc, umi, tag numbering system (refer to kallisto documentation).
  • In the general settings:
    • run_modes section will toggle different modalities to be run.
    • threads and memory allow rule-specific thread/memory setting respectively. This pipeline is designed for LSF based submission but the lsf.yml file can be edited as needed (see below)
paths:
  atac_fastq_dir: ""
  cellranger_ref_dir: ""
  cellranger_exe: "/home/wuv/pkg/cellranger-atac-2.1.0/bin/cellranger-atac"
  adt_fastq_dir: ""
  adt_catalog: "adt_panel.csv"
  allowlist: "737K-cratac-v1.revcomp.txt"
  amulet_dir: "AMULET"
  autosomes: "hg38_autosomes.txt"
  denylist: "hglft_genome_411bb_f9b580.bed"


out_dirs:
  atac: "." # this currently cannot be changed
  haystack: "haystack_out"
  adt: "adt_out"
  amulet: "amulet_out"

general:
  run_modes:
    atac: True
    atac_level: "namesort"
    adt: True

  library:
    cbc: [1,0,16]
    umi: [0,0,10]
    tag: [2,0,15]

  threads:
    cellranger_count: 12
    namesort: 8
    make_bus: 4
    sort_bus: 8
    amulet: 2

  memory:
    cellranger_count: 120000
    namesort: 96000
    make_bus: 36000
    amulet: 36000

samples:
  - name: "A"
    adt_fastqs: ["A_S2_R1_001.fastq.gz", "A_S2_R2_001.fastq.gz", "A_S2_R3_001.fastq.gz"]

2) lsf.yaml

This file allows for specific LSF submission settings based on rules.

__default__:
  - "-q normal"


cellranger_count:
  - "-q denovo"

namesort:
  - "-q denovo"

3) fastq files for ADT; directory of fastq files for ATAC

Refer to config above

4) Antibody catalog csv file (required if doing ADT portion)

Comma separted file with barcodes and descriptions. Example below:

CD40,CTCAGATGGAGTATG
CD44,AATCCTTCCGAATGT
CD48,CTACGACGTAGAAGA
CD21,AACCTAGTAGTTCGG

4) Allowlist cell barcode file (required if doing ADT portion)

Allowlist file containing accepted cell barcodes. One per line. Example below:

GTCTGCTATGTCTA
GATGATGCATAGAA

Setup

  1. Create conda env and set up LSF job submission for Snakemake (follow instructions here).
conda create --name <env> --file env.yml
  1. Run Snakemake
bsub -e snek.e -o snek.o snakemake --configfile=config.yaml --profile=lsf -s <path-to-Snakefile>

The overall process will look like this (to be added):

  1. Downstream analysis of your own choosing. Output files are as follows (to be added)...

Credits