Update to CNV workflow

[sirselim]

I had some time recently for another job to better implement the CNV module/workflow. It's still not great, the main branch needs some more love, but the below should get you 99% of the way towards working code.

#!/bin/bash -l

# check and grab latest wf (if wanted)
nextflow pull epi2me-labs/wf-cnv

# define variables
WKDIR='/data/basecalled/AGRF'
SAMPLE='Diabetes_A'
REFERENCE='/public-data/references/GRCh38/GCA_000001405.15_GRCh38_no_alt_analysis_set.fna'
NFCONFIG='/public-data/configs/nextflow_local_overide.config'
THREADS='48'

# EPI2ME Labs CNV workflow
# nextflow -c "${NFCONFIG}" run epi2me-labs/wf-cnv \
#   -resume \
#   --threads "$THREADS" \
#   -profile standard,local \
#   --fastq "${WKDIR}"/"${SAMPLE}"/pass/ \
#   --sample "${SAMPLE}" \
#   --fasta "${REFERENCE}" \
#   --genome hg38 --bin_size 500 \
#   --map_threads 24


# EPI2ME Labs CNV workflow
nextflow -c "${NFCONFIG}" run epi2me-labs/wf-cnv \
  -resume \
  --threads "$THREADS" \
  -profile standard,local \
  --fastq "${WKDIR}"/"${SAMPLE}"/pass/fastq/ \
  --sample_sheet sample_sheet.csv \
  --fasta "${REFERENCE}" \
  --genome hg38 --bin_size 500 \
  --map_threads 24


# Notes:
# EPI2ME Labs workflow for ONT CNV analysis: https://github.com/epi2me-labs/wf-cnv
#  CNV pipeline will run on the basecalled fastq files, it will run an alignment against 
# the provided reference genome, and then perform CNV analysis using QDNAseq.

# there is currently a fun little issue with the CNV pipeline needing a sample sheet
# it also seems to ignore non-compressed fastq files, so run:
# for file in *.fastq ; do 
#   echo -e "... processing $file ...";
#   bgzip "$file";
#   echo -e "... done ...";
# done
# NOTE: it also seems that the gz fastq files have to be in a folder called fastq...

# sample sheet looks like this:
# barcode,sample_id,type
# barcode01,Diabetes_A,AGRF_Diabetes

[leahkemp]

Wicked thanks for sharing your code Miles, I've implemented it and will test/refine it when I get to that section of the code :)

[leahkemp]

5daf211

[sirselim]

Reopening this due to a big update in the wf-cnv workflow. You no longer are required to use a samplesheet or just fastqs. Bams are now a valid entry point which massively speeds up the CNV calling process. Some example code:

#!/bin/bash -l

# check and grab latest wf (if wanted)
nextflow pull epi2me-labs/wf-cnv

# define variables
WKDIR='/data/basecalled/AGRF'
SAMPLE='Control'
REFERENCE='/public-data/references/GRCh38/GCA_000001405.15_GRCh38_no_alt_analysis_set.fna'
NFCONFIG='/public-data/configs/nextflow_local_overide.config'
THREADS='48'


# EPI2ME Labs CNV workflow
nextflow -c "${NFCONFIG}" run epi2me-labs/wf-cnv \
  -resume \
  --threads "$THREADS" \
  -profile standard,local \
  --bam "${WKDIR}"/"${SAMPLE}"/bam/"${SAMPLE}"_sorted_merged.hp.bam \
  --sample "${SAMPLE}" \
  --fasta "${REFERENCE}" \
  --genome hg38 --bin_size 500 \
  --map_threads 24


# Notes:
# EPI2ME Labs workflow for ONT CNV analysis: https://github.com/epi2me-labs/wf-cnv
# CNV pipeline will run on the basecalled fastq files or aligned bams, 
# and then perform CNV analysis using QDNAseq.