how to copy and rename files in nextflow

Question

[fabiomarcelomedvet]

Hi everyone. I've been trying to copy and rename files created by a pipeline executed in nextflow. I have implemented a qiime2 workflow and got the taxonomy (.qza) and biom files (see image below). Wen I run qiime tools export qiime2 creates a directory containing the tsv and/or biom file . The entire pipeline is executed in container = quay.io/qiime2/core:2023.7 with runOptions = -u $(id -u):$(id -g).

The process which exports the taxonomy is the following:

process taxonomy{
  publishDir params.outdir, mode:'copy'

  input:
  path "table-denoised.qza"
  path "vsearch-taxonomyITS.qza"
  path "blast-taxonomyITS.qza"
  path "sklearn-taxonomyITS.qza"

  output:
  path "feature-table", emit: feature_table
  path "vsearch-taxonomy", emit: vsearch_taxonomy
  path "blast-taxonomy", emit: blast_taxonomy
  path "sklearn-taxonomy", emit: sklearn_taxonomy

  script:
  """
  qiime tools export \
  --input-path table-denoised.qza \
  --output-path feature-table

  qiime tools export \
  --input-path vsearch-taxonomyITS.qza \
  --output-path vsearch-taxonomy

  qiime tools export \
  --input-path blast-taxonomyITS.qza \
  --output-path blast-taxonomy

  qiime tools export \
  --input-path sklearn-taxonomyITS.qza \
  --output-path sklearn-taxonomy
  """
}

Now I need to copy and rename the tsv files. I've tried:

process replace_header{
  publishDir params.outdir, mode:'copy'

  input:
  path "vsearch-taxonomy/taxonomy.tsv"

  output:
  path "taxonomy.tsv"

  script:
  """
  cp vsearch-taxonomy/taxonomy.tsv taxonomy.tsv
  """
}

Any idea?

[Pierre Lindenbaum]

please validate/comment the answers to your previous questions: How to install non conda software ; Age nodes in R

[Ram]

Please use the formatting bar (especially the code option) to present your post better. You can use backticks for inline code (`text` becomes text), or select a chunk of text and use the highlighted button to format it as a code block. If your code has long lines with a single command, break those lines into multiple lines with proper escape sequences so they're easier to read and still run when copy-pasted. I've done it for you this time.

Answer 1

[fabiomarcelomedvet]

Solution:

1) Add output name as dir/filename after qualifier in the process taxonomy:

`process taxonomy{
  publishDir params.outdir, mode:'copy'

  input:    
  file "table-denoised.qza"    
  file "sklearn-taxonomyITS.qza"    
  file "vsearch-taxonomyITS.qza"    
  file "blast-taxonomyITS.qza"

  output:
  path ("feature-table/feature-table.biom"), emit: feature_table_biom
  path ("sklearn-taxonomy/taxonomy.tsv"), emit: sklearn_taxonomy_tsv
  path ("vsearch-taxonomy/taxonomy.tsv"), emit: vsearch_taxonomy_tsv
  path ("blast-taxonomy/taxonomy.tsv"), emit: blast_taxonomy_tsv

  script:    
  """
  qiime tools export \
  --input-path table-denoised.qza \
  --output-path feature-table

  qiime tools export \
  --input-path sklearn-taxonomyITS.qza \
  --output-path sklearn-taxonomy

  qiime tools export \
  --input-path vsearch-taxonomyITS.qza \
  --output-path vsearch-taxonomy

  qiime tools export \
  --input-path blast-taxonomyITS.qza \
  --output-path blast-taxonomy
  """
} 

`

2) process replace_header: define the paths in output chunck as the names of the files; In the script chunck provide full path to the file to be edited and redirect the output into a new file.

process REPLACE_HEADER {
  publishDir params.outdir, mode:'copy'

  input:
  path "vsearch-taxonomy/taxonomy.tsv"
  path "sklearn-taxonomy/taxonomy.tsv"
  path "blast-taxonomy/taxonomy.tsv"

  output:
  path ("vsearch_taxonomy.tsv"), emit: file_vsearch_taxonomy_tsv
  path ("sklearn_taxonomy.tsv"), emit: file_sklearn_taxonomy_tsv
  path ("blast_taxonomy.tsv"), emit: file_blast_taxonomy_tsv

  script:
  """
  sed 's/Feature ID/#otu-id/g' vsearch-taxonomy/taxonomy.tsv > vsearch_taxonomy.tsv
  sed 's/Feature ID/#otu-id/g' sklearn-taxonomy/taxonomy.tsv > sklearn_taxonomy.tsv
  sed 's/Feature ID/#otu-id/g' blast-taxonomy/taxonomy.tsv > blast_taxonomy.tsv
  """
}

Answer 2

[Pierre Lindenbaum]

unless I'm wrong your first process could contain the following output:

  path ("vsearch-taxonomy/taxonomy.tsv"), emit: vsearch_taxonomy_tsv

then, you want to rename/move this output. This is probably NOT what you really want to do. Just use the named output 'vsearch_taxonomy_tsv` and reuse it in another place.

you can always rename it at the end using publish

[fabiomarcelomedvet]

Thanks for replying, Pierre!

I've tried your sugestion but when I try to copy (or whatever action) the tsv file I got No such file or directory.

Every qiime tools export in the process taxonomy results a folder (defined by the paramerter --output-path) inside which is the tsv file, as you can see below. There's no way to save the file in the current folder.

Answer 3

[pbioinf]

Use the saveAs parameter from the publishDir directive to save the outputfile with a specific name.

https://www.nextflow.io/docs/latest/process.html#publishdir

Untested:

publishDir params.outdir, mode:'copy', saveAs {it -> 'taxonomy.tsv'}

[fabiomarcelomedvet]

Thanks for replying. I got the error:

N E X T F L O W ~ version 23.10.0 Launching../github/its_pipeline/main.nf` [soggy_wright] DSL2 - revision: 2b94ede702 ERROR ~ Script compilation error

• file : /home/mapa/github/its_pipeline/main.nf
• cause: The current parameter list already contains a parameter of the name it @ line 390, column 49. s.outdir, mode:'copy', saveAs {it -> 'ta

                             ^
  

1 error

-- Check '.nextflow.log' file for details`