gghist.R does not exist in cbcrg/callings-with-gatk:latest

[Xophmeister]

Step 6B in the RNA-seq variant calling pipeline work-through uses the DockerHub container cbcrg/callings-with-gatk:latest. This image exists, but it does not contain the executable script gghist.R, which is used in the process script. Thus the workflow fails.

Code

training/hands-on/final_main.nf

Lines 309 to 338 in d46bd35

	process prepare_vcf_for_ase {
	container 'cbcrg/callings-with-gatk:latest'
	tag "${sampleId}"
	publishDir "${params.results}/${sampleId}"
	input:
	tuple val(sampleId),
	path('final.vcf'),
	path('commonSNPs.diff.sites_in_files')
	output:
	tuple val(sampleId), path('known_snps.vcf'), emit: vcf_for_ASE
	path 'AF.histogram.pdf' , emit: gghist_pdfs
	script:
	'''
	awk 'BEGIN{OFS="\t"} $4~/B/{print $1,$2,$3}' commonSNPs.diff.sites_in_files > test.bed
	vcftools --vcf final.vcf --bed test.bed --recode --keep-INFO-all \
	--stdout > known_snps.vcf
	grep -v '#' known_snps.vcf | awk -F '\\t' '{print $10}' \
	| awk -F ':' '{print $2}' | perl -ne 'chomp($_); \
	@v=split(/\\,/,$_); if($v[0]!=0 ||$v[1] !=0)\
	{print $v[1]/($v[1]+$v[0])."\\n"; }' | awk '$1!=1' \
	> AF.4R
	gghist.R -i AF.4R -o AF.histogram.pdf
	'''
	}

Documentation (first mention; repeated elsewhere in this file)

training/docs/hands_on/04_implementation.md

Lines 1346 to 1376 in d46bd35

	process prepare_vcf_for_ase {
	container 'cbcrg/callings-with-gatk:latest'
	tag "${sampleId}"
	publishDir "${params.results}/${sampleId}"
	input:
	tuple val(sampleId),
	path('final.vcf'),
	path('commonSNPs.diff.sites_in_files')
	output:
	tuple val(sampleId), path('known_snps.vcf'), emit: vcf_for_ASE
	path 'AF.histogram.pdf' , emit: gghist_pdfs
	script:
	'''
	awk 'BEGIN{OFS="\t"} $4~/B/{print $1,$2,$3}' \
	commonSNPs.diff.sites_in_files > test.bed
	vcftools --vcf final.vcf --bed test.bed --recode --keep-INFO-all \
	--stdout > known_snps.vcf
	grep -v '#' known_snps.vcf | awk -F '\\t' '{print $10}' \
	| awk -F ':' '{print $2}' | perl -ne 'chomp($_); \
	@v=split(/\\,/,$_); if($v[0]!=0 ||$v[1] !=0)\
	{print $v[1]/($v[1]+$v[0])."\\n"; }' | awk '$1!=1' \
	>AF.4R
	gghist.R -i AF.4R -o AF.histogram.pdf
	'''
	}

[Xophmeister]

...Ah, my mistake, this script should be in the workflow's bin directory. It's briefly mentioned in the setup chapter:

training/docs/hands_on/03_setup.md

Lines 23 to 48 in d46bd35

	```bash
	/workspace/gitpod/hands-on
	hands-on
	├── README.md
	├── bin
	│   └── gghist.R
	├── data
	│   ├── blacklist.bed
	│   ├── genome.fa
	│   ├── known_variants.vcf.gz
	│   └── reads
	│   ├── ENCSR000COQ1_1.fastq.gz
	│   ├── ENCSR000COQ1_2.fastq.gz
	│   ├── ENCSR000COQ2_1.fastq.gz
	│   ├── ENCSR000COQ2_2.fastq.gz
	│   ├── ENCSR000COR1_1.fastq.gz
	│   ├── ENCSR000COR1_2.fastq.gz
	│   ├── ENCSR000COR2_1.fastq.gz
	│   ├── ENCSR000COR2_2.fastq.gz
	│   ├── ENCSR000CPO1_1.fastq.gz
	│   ├── ENCSR000CPO1_2.fastq.gz
	│   ├── ENCSR000CPO2_1.fastq.gz
	│   └── ENCSR000CPO2_2.fastq.gz
	├── final_main.nf
	└── nextflow.config
	```

Maybe this could be made clearer?

[mribeirodantas]

You're correct. It should be made more explicit. Thanks, @Xophmeister! 😉

Do you want to have a try with a PR?