lbeckman314
commented
5 months ago It looks like the tuple is being treated as a single parameter, as opposed to 5 separate parameters.
We can update the input section to handle this with the tuple qualifier:
// Define the Mutect2 process
process mutect2 {
...
- input:
- val chrom
- path mutect_idx
- path tumor_bam_sorted
- path normal_bam_sorted
- path pon
+ tuple val(chrom),
+ path(mutect_idx),
+ path(tumor_bam_sorted),
+ path(normal_bam_sorted),
+ path(pon)Can you try re-running the command with this change (included in the 'Full Script' below)?
Full Script
```nf #!/usr/bin/env nextflow nextflow.enable.dsl=2 // Define the list of chromosomes chromosomes = (1..22).collect { it.toString() } + ['X'] // Create a channel emitting each chromosome chrom_channel = Channel.from(chromosomes) // Define the Mutect2 process process mutect2 { memory '80 G' publishDir "${params.outdir}/svc", mode: 'copy' input: // Deconstruct the tuple into separate inputs tuple val(chrom), path(mutect_idx), path(tumor_bam_sorted), path(normal_bam_sorted), path(pon) output: path("${params.id}_chr${chrom}_unfiltered.vcf") path("${params.id}_chr${chrom}_f1r2.tar.gz") path("${params.id}_chr${chrom}_unfiltered.vcf.stats") script: """ gatk Mutect2 \\ -R ${mutect_idx} \\ -I ${tumor_bam_sorted} \\ -I ${normal_bam_sorted} \\ --panel-of-normals ${pon} \\ -normal ${normal_bam_sorted.baseName} \\ -L ${chrom} \\ -O ${params.id}_chr${chrom}_unfiltered.vcf \\ --f1r2-tar-gz ${params.id}_chr${chrom}_f1r2.tar.gz \\ -stats ${params.id}_chr${chrom}_unfiltered.vcf.stats """ } // Run the process for each chromosome workflow { chrom_channel .map { chrom -> tuple(chrom, params.mutect_idx, params.tumor_bam_sorted, params.normal_bam_sorted, params.pon) } | mutect2 } ```
elisabethgoldman
rlancaster96
Problem:
Problem in the way I'm using input and/or output channels to pass multiple chromosomes to Mutect2 for per-chromosome processing; leads to
Error: Process mutect2 declares 5 input channels but 1 were specified. Suspect both the channel and workflow setup have errors; there are several ways to set up multiple dynamically-named output files, but none I have tried so far have succeeded. Plan is to merge the 23 files with bcftools and GATK4 commands subsequently (those are largely figured out). See the following for output file options: (https://www.nextflow.io/docs/latest/process.html#multiple-output-files)Desired Behavior:
Split mutect2 processing by chromosome for speed, and thus output the 3 output file types by chromosome
Desired Output:
Output 23 files for each of three file types (unfiltered.vcf, f1r2.tar.gz, stats).
sample1_chr1.unfiltered.vcf,...., sample1_chrX.unfiltered.vcf)sample1_chr1.f1r2.tar.gz...)sample1_chr1.stats...)Nextflow process and subworkflow (pasted together for issue but modularized in repo)
params-file
Error log: