MultiQC not displaying all elements in a table

Hello there lovely people - I’m hoping you might be able to assist me

I have a table of partial AMR matches arranged in a table and I want to include these in a MultiQC report. There may be multiple partial detections for some specimens - see the snip directly below.

When the MultiQC report is generated in galaxy (for the table I have specified custom content > table), it appears that MultiQC is only displaying a single line for each specimen, although there may be multiple partial matches in the corresponding table that should also be reported on. I’ve read some threads that mention issues around samples with the same name, but this is a single table and I’m unclear if this should even be a problem in that respect. I’ve had a read of the MultiQC documentation and I can’t find anything that sufficiently addresses this issue.

I wonder if you might be able to assist me.

Thanks, Graham

Welcome @scyllas1981

This is a good question!

Yes, MultiQC does some data massaging around sample keys. This happens wherever it is used (not just in Galaxy).

The tool expects that the first column is always a unique index key, the sample identifier. (It can also remove certain file extensions).

I can think of a few ways to get all lines displayed, and the easiest could be simply adding in an “index” column as the first column. The tools used could be:

  • Add column** to an existing dataset (with the Iterate? toggled to YES)
  • Then rearrange the columns with Cut columns from a table

You could also make the column more meaningful. The important part is that the key is unique. If you are new to Text Manipulation in Galaxy, these tutorials can help you to get oriented. Not all tools are included but you can usually search with common utility names then find help on the tool form directly. Please ask if you have questions!

I hope this helps and we can follow up more! :slight_smile:

Hi @jennaj - thanks for your response!

I had a feeling this might be the case; I understand the logic behind the limitations (and creating a new unique key is absolutely no issue).

Introducing another process to create a new unique key just to display data in this format is likely unnecessary and I’m happy for it to just be a plain tsv output in the end.

Would be nice though if it were a little more flexible in this regard.

Thanks for your help

Graham

I agree! I wish there was a little flag to tell the tool “just print this out!” and I think the source repository hosted from MultiQC has issue tickets about this, too, but I don’t know the status or future development plans.

Hum, let;s check → GitHub - MultiQC/MultiQC: Aggregate results from bioinformatics analyses across many samples into a single report.

I don’t see anything obvious. Maybe I just saw that discussion at other forums at some point.

But – glad you have this working!