select a specific output file from one process and use an an input to another process in nextflow

[awast]

I am writing a nextflow based script for gatk based analysis on normal-tumor paired samples.
In the below mentioned process, i got multiple output files but for running next process i need only one file from it i.e.(recal_readsbamfile)

process gatkrun {
      
      publishDir params.trim, mode : 'copy' 

      input:
      tuple val(pair_id), file(sorted_file)

      output:
      tuple val(pair_id),file(alignment_metrics_file),file(insert_size_metrics_file),file(Histogram_file),file(readgroup_bamfile),file(dedup_readsbamfile),file(dedup_readsbaifile),
      file(dedup_metricsfile),file(recal_datatable),file(recal_readsbamfile),file(post_recal_datatable),file(recalibration_plots)
     
      script:
      alignment_metrics_file = pair_id + '_alignment_metrics.txt'
      insert_size_metrics_file = pair_id + '_insertsize_metrics.txt'
      Histogram_file = pair_id + '_insert_size_histogram.pdf'
      readgroup_bamfile = pair_id + '_readgroup.bam'
      dedup_readsbamfile = pair_id + '_dedup_reads.bam'
      dedup_readsbaifile = pair_id + '_dedup_reads.bai'
      dedup_metricsfile = pair_id + '_dedup_metrics.txt'
      recal_datatable = pair_id + '_recal_data.table'
      recal_readsbamfile = pair_id + '_recal_reads.bam'
      post_recal_datatable = pair_id + '_postrecal_data.table'
      recalibration_plots = pair_id + '_recalibration_plots.pdf'
       
      
     """ 
     ./gatk  CollectAlignmentSummaryMetrics -R ${params.reference} -I $sorted_file -O $alignment_metrics_file 
     /./gatk CollectInsertSizeMetrics -I $sorted_file -O $insert_size_metrics_file -H $Histogram_file
     ./gatk  AddOrReplaceReadGroups --INPUT $sorted_file  --OUTPUT $readgroup_bamfile --RGLB lib1 --RGPL illumina --RGPU NONE --RGSM $pair_id
   ./gatk  MarkDuplicates -I $readgroup_bamfile -O $dedup_readsbamfile -M $dedup_metricsfile -AS true --VALIDATION_STRINGENCY LENIENT
    ./gatk  BuildBamIndex --INPUT $dedup_readsbamfile --OUTPUT $dedup_readsbaifile 
    ./gatk - BaseRecalibrator -R ${params.reference} -I $dedup_readsbamfile --known-sites ${params.snpfile} -O $recal_datatable
    ./gatk  ApplyBQSR -R ${params.reference} -I $dedup_readsbamfile -bqsr $recal_datatable -O $recal_readsbamfile 
    ./gatk  BaseRecalibrator -R ${params.reference} -I $recal_readsbamfile --known-sites ${params.snpfile} -O $post_recal_datatable
    ./gatk -AnalyzeCovariates -before $recal_datatable -after $post_recal_datatable -plots $recalibration_plots
     """

}

in the next process i want to give input of only one file from above process but it takes the very first output file as an input in next process.
Please help me in resolving this issue

[bentsherman]

Hi @awast , I would recommend that you split your output declaration into multiple outputs:

      output:
      tuple val(pair_id), file(alignment_metrics_file)
      tuple val(pair_id), file(insert_size_metrics_file)
      // ...
      tuple val(pair_id), file(recal_readsbamfile), emit: recal_readsbamfile

Then, when you call this process in the workflow, you can pass individual outputs as you need them:

workflow {
    gatkrun(ch_input)

    next_process(gatkrun.out.recal_readsbamfile)
}

Going even further, it might be good to split each of these GATK commands into separate processes. That way Nextflow can parallelize these steps as much as possible.

[awast]

Thanks for the quick response.
I have included the above mentioned changes and its working.
Thanks again