mirtrace casting error - with a fix

[seiboldfyr]

Description of the bug

Our pipeline errors out with a casting error when entering mirtrace in version 2.3 (and dev version) of the smRNAseq pipeline. It works fine with only fastp adapter trimming, but when we use the UMI tools and fastp we consistently get the java type casting bug.

The fix is: edit the smrnaseq/modules/local/mirtrace.nf file, line 10. The pipeline uses path(reads), but when that's changed to val(reads) the pipeline completes without the casting bug. Verifies that it removes UMIs and adapters as expected as well.

Command used and terminal output

input: 

nextflow run nf-core/smrnaseq -profile docker -r dev -c ${inputConfig} --input ${inputFile} --outdir '.' -work-dir './work/' --genome GRCh38 --mirtrace_species 'hsa' --with_umi true --umitools_extract_method regex --umitools_bc_pattern '.+(?P<discard_1>AACTGTAGGCACCATCAAT){s<=2}(?P<umi_1>.{12})(?P<discard_2>.*)' --save_umi_intermeds true --protocol 'qiaseq' --fastp_min_length 15


output: 

ERROR ~ Error executing process > 'NFCORE_SMRNASEQ:MIRTRACE:MIRTRACE_RUN (1)'

Caused by:
  Cannot cast object 'qiaseq_Adeptlinear.fastp.fastq.gz' with class 'nextflow.processor.TaskPath' to class 'java.util.List' -- Check script '/home/fyrseq/.nextflow/assets/nf-core/smrnaseq/./workflows/../subworkflows/local/../../modules/local/mirtrace.nf' at line: 28

Source block:
  def primer = adapter ? "--adapter ${adapter}" : ""
  def protocol = params.protocol == 'custom' ? '' : "--protocol $params.protocol"
  def java_mem = ''
  if(task.memory){
          tmem = task.memory.toBytes()
          java_mem = "-Xms${tmem} -Xmx${tmem}"
      }
  def config_lines = [ids,reads]
      .transpose()
      .collect({ id, path -> "echo '${path},${id}' >> mirtrace_config" })

Relevant files

nextflow.log

System information

• smRNAseq version dev (also tested with 2.3.0)
• Nextflow version: 23.10.1.5891
• Java version: openjdk version "11.0.22" 2024-01-16
• Operating system: Linux
• Bash version: GNU bash, version 5.0.17(1)-release (x86_64-pc-linux-gnu)

[apeltzer]

smrnaseq/modules/local/mirtrace.nf

Line 10 in c611b15

tuple val(adapter), val(ids), path(reads)

I can't find the issue here...?

[seiboldfyr]

Ah sorry, I switched the words! Meant to say that path(reads) errors, but val(reads) works. Just corrected my post.

[apeltzer]

Does that then work for multiple samples too?

[seiboldfyr]

Yes, we ran it on a set of 4 Qiaseq samples

[apeltzer]

Please run with -r dev or soon -r 2.3.1 - this should be fine then