Troubleshooting: Krona Pie chart display

Hi. I am using galaxy at https://usegalaxy.org/
I have a similar problem that was discussed above.
I am working on a metagenomics project and trying to use Krona pie chart tool to visualize my data. In recent runs, it stopped generating the chart. Instead, it began to display hypertext coding for the HTML.

The results appear in green but it says the following:
“”“!DOCTYPE html PUBLIC “-//W3C//DTD XHTML 1.0 Strict//EN” “http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd
html xmlns=“XHTML namespace” xml:lang=“en” lang=“en”
head
meta charset=“utf-8”/”“”

what should I do in order to visualize our results file?
Thanks

Welcome, @vaandio

That issue from 6 years ago was about the HTML display of an output file needing a flag changed in the database, and has long been resolved.

Your display message seems like it could be something similar, or there is some other problem with the file.

There are a few versions of these tools, and I just checked at UseGalaxy.org and all have the setting display technically correct. Maybe the problem was transient – have you tried again yet?

If this is still presenting, we can look at the exact tool and version are you using. You can share your work to help clarify this, how-to is in the banner of the forum, also here. → How to get faster help with your question

Let’s start there, and let us know if this has resolved already! The server had some technical changes underway the last few days, and trying again is the usual solution for small display issues but perhaps not all. We can follow up to confirm but will need an example. :slight_smile:

Dear jennaj!

I tried to reload the KRONA results file, but nothing changed.
I’ll share with you the history:
Galaxy

Thank you for your help!
BEst regards
vaandio

On Tue, 21 Jan 2025 20:40:52 +0000, Jennifer Hillman-Jackson via Galaxy Community Help wrote

| jennaj
> January 21 |

  • | - |

Welcome, @vaandio
That issue from 6 years ago was about the HTML display of an output file needing a flag changed in the database, and has long been resolved.
Your display message seems like it could be something similar, or there is some other problem with the file.
There are a few versions of these tools, and I just checked at UseGalaxy.org and all have the setting display technically correct. Maybe the problem was transient – have you tried again yet?
If this is still presenting, we can look at the exact tool and version are you using. You can share your work to help clarify this, how-to is in the banner of the forum, also here. → How to get faster help with your question
Let’s start there, and let us know if this has resolved already! The server had some technical changes underway the last few days, and trying again is the usual solution for small display issues but perhaps not all. We can follow up to confirm but will need an example. Embedded CGI removed by webmail. https://emoji.discourse-cdn.com/twitter/slight_smile.png?v=12

Hi @vaandio

The tool itself is displaying data in general, it is the content of your files that are the root problem. The file we are looking at in your history is about 4-5x the size of what is usually expected. The content itself is fine. These are just interactive HTML files, so the actual numbers (numbers representing the counts) on rows is not going to make the file larger, but the number of species (rows) definitely could.

To see what is in your HTML file, you can download it, then uncompress and open in any web browser from your own computer. Now, I tried this with just one of your files run through Krona, not both together, but you can try that.

Notice how the graphic is mostly a black box on one side due to the density of the lower count calls and Krona’s attempt to render them all. You could try using more stringent parameters with Kraken2, or post filtering the data in other ways to focus in with more resolution in branches of interest. I tested this at a few of the different public Galaxy servers. All timed out with the web browser display. Our developers think that is expected given the file size.

Hope this helps to explain what is going on, and provides an alternative! :slight_smile:

Dear Jennaj,

Thank you for your advice. I tried to dowload and open the html file, but nothing was seen in the webbrowser.

Could the origin of the problem been by computer and its powersources?
I attach a printscreen about my pc parameters and the version number of the used webbrowser (Microsoft Edge)

Thanks for your further help and instruction.

BEst regards
Andrea

admin edit reformat

Hi @vaandio

None of this should be dependent on your computer specifications. I don’t use PCs but you can probably translate these MacOS instructions for a PC, since this is just uncompressing the file to plain HTML then using a web browser.

  1. Download the file from Galaxy → FAQ: Downloading datasets
  2. Uncompress from html.gz to html. On a MacOS I just double click and that happens but that is the OS doing it automatically (maybe I set a preference, can’t recall). You could use any tool you would normally use to get a gzip’d file out into plain text. Try a web search with those words and you’ll probably find instructions.
  3. Then from your web browser, try to open the local plain text html file.

I was able to do this with Google Chrome, which should be also available on a PC. I also only tested this with a rerun of just one of your input datasets (the smaller one) and not the larger one, and not both together. So maybe try that too just to see what this is looking like (not very useful).

For the root issue: you have raw outputs that are not immediately useful for a graphic. You could explore the stringency of those calls: confidence, base quality, kmer overlap, and others. Some of this you can do with the other post-processing tools; search the tool panel with “kraken” to find these. Most have a tutorial plus you can just experiment. Maybe see if a publication is doing the same analysis: exploring a large complex sample. What tools did they use? Galaxy might have those or something similar. :scientist:

Also, remember that the GTN is not the only source of tutorials. As an example – using just a web browser general search with “how to view very large kraken reports in krona” – this was the second hit (after another hit back to this forum).

  • Metagenomics walkthrough - CLIMB-BIG-DATA Documentation
    Everything done here can be done in Galaxy and some is in our tutorials, and some not exactly because there are so many ways to explore! The tools and the RStudio environment are the same – the manipulations and ideas differ. Search the tool panel to find tools including BLAST and any Galaxy specific tutorials will be linked on tool forms (to help understand interactive environments and how to move your data in and out of those from a Galaxy history).

Hope this helps! You have lots of choices about what to do next.