Hi everyone , I'm currently working on the development of a pipeline to process the BLISS ( break labelling in situ sequencing) data, but I have a bug to fix. I launch it with as input a csv of 2 samples paired-end, it finish without error but losing one sample and then when I relaunch it with -resume , the lost sample is process without error. At the beggining of the pipeline I used the subworkflow fastq_fastqc_umitools_trimgalore , then bwa_mem and umitools_dedup and it seems that the lost of one of two samples Can someone help me ?

If a process is supposed to run multiple times but spawns only a single task, then in 99% of the cases the problem is caused by the consumption of a single-element queue channel that is supposed to be a value channel, e.g. the reference genome channel.

thanks !