-
Hi, I'm new to NextFlow so forgive me if I've missed something very basic here. I'm trying to build a pipeline with the following steps:
I've tried to run the pipeline. All steps work except the final MultiQC process where I get an error:
However, checking the output folders - there are no duplicate file names. Furthermore, I've tried to avoid any collision by naming prefixing the 1M subsampled FASTQ files with "1M-" - so they do have distinct names from the post-trimming FASTQ files. Any help would be much appreciated! My pipeline:
|
Beta Was this translation helpful? Give feedback.
Replies: 7 comments
-
I would look at how the nf-core pipelines do it. For example it looks like nf-core/rnaseq defines separate inputs for the MULTIQC process: https://github.com/nf-core/rnaseq/blob/master/modules/local/multiqc/main.nf |
Beta Was this translation helpful? Give feedback.
-
Thanks for the suggestions. I tried defining separate inputs for the MULTIQC process as suggested the two inputs were the two directories where the two FASTQC process outputs were saved - however I'm still getting an error:
As before, the files in each output folder are unique and the two directory names are unique. I also tried going back to the mix and collect channels approach and get the same error. My current current script is:
Any help would be much appreciated. |
Beta Was this translation helpful? Give feedback.
-
check response by muffato, Matthieu Muffato |
Beta Was this translation helpful? Give feedback.
-
Thanks, but I don't understand why this should be necessary - the outputs are in two different sub directories and are all named uniquely. This is the file list:
|
Beta Was this translation helpful? Give feedback.
-
@jamesboot Sorry, I do not know. |
Beta Was this translation helpful? Give feedback.
-
I think when you have a single Improving this case is something I can look into as part of #4553 |
Beta Was this translation helpful? Give feedback.
-
Hi, thanks for your suggestions, in the end I changed the FASTQC processes to output a file instead of a path using:
Instead of:
Then in the workflow I defined the input for the MULTIQC process as the collection of the two channels coming out of the FASTQC processes:
Finally I defined the two inputs separately in the MULTIQC process:
This has fixed the problems. Thanks for the suggestions above. |
Beta Was this translation helpful? Give feedback.
Hi, thanks for your suggestions, in the end I changed the FASTQC processes to output a file instead of a path using:
Instead of:
Then in the workflow I defined the input for the MULTIQC process as the collection of the two channels coming out of the FASTQC processes:
Finally I defined the two inputs separately in the MULTIQC process:
This has fixed the problems. Thanks for the suggestions above.