Addition of several plots to MultiQC

[jennaj]

Hi @Orange_Pomeranian

For this

@Orange_Pomeranian

Thanks for your reply. I tried connecting the inputs to the ‘Merge collections’. Sadly most of the inputs don’t work (even when I remove downstream connections).

You can try a few other things:

1. Disconnect all noodles, then reconnect in the order of job execution
2. Double check that you are merging collections that contain data all with the same datatype. For your case, this would be the txt formatted reports for MultiQC: one merge per report type
3. Then, that merged collection is connected to MultiQC, after the section is defined to be a collection and the tool/report type is set

Then for this

@Orange_Pomeranian

I want to merge several plots into one collection so that it can be added to the final multiqc to the Image Content section. Merge collections seems to don’t work with that. Do you have any example workflow where its shown how it can be connected?

This tutorial has a general merge example

• Hands-on: Using dataset collections / Using dataset collections / Using Galaxy and Managing your Data (merge-collections)

And this workflow template includes MultiQC with a merge in the data preparation steps.

• Quality Control Start Here! multQC issue and guidance? - #2 by jennaj << see the workflow template
• Issues with receiving results from FastQC in MultiQC from several collections of samples - #3 by jennaj << screenshots and direct link to the workflow

Notice how I was ensuring that all samples had a unique element identifier (sample name) at the very start. Paired fastq data is a bit different from BAMs and Bigwig data, and involves an extra step to add in forward/reverse notation inside of the workflow, but the basic logic applies to all data: simple, standardize formats for the element identifiers used in the input collections. This makes it easier to manipulate the collections later.

Finally, I have some unsolicited advice : your workflow contains duplicated processing paths. Could this be simplified? Maybe input all BAMs together, all BigWigs together, run the statistics together, then split out control/case data after into collections (using the element identifiers from a tabular input?) for the downstream MACS2 jobs?

But either way should work. You can welcome to share back a link to your workflow (and screenshots showing what to pay attention to) as you work through this. But try the first items above first please! Thanks!