Error with opening Metadata

[bbosnja]

Hi,
While trying to run the script:
marmot(metadata = "D:/For MARMOT/MARMOT_Metadata.xlsx",
name = "LF vs HF",
render = TRUE
)

I got the following error:
"Fehler in openxlsx::getSheetNames(metadata) : file does not exist."

The same file can be opened without any problems upon running:
openxlsx::getSheetNames("D:/For MARMOT/MARMOT_Metadata.xlsx") :
[1] "Pipeline Settings" "Study Data" "File Data" "Options"

I would appreciate your help with this issue.
Best,
Berislav

[peterleary]

Hi Berislav,

Thanks for giving marmot a try! And apologies for the issue. The code was only looking in C:/ drive on Windows, so I've pushed a quick fix to let it load from any lettered drive, including D! Please reinstall the R package, restart R, and try again, it should work now.

The only caveat is, I don't have access to a Windows machine, so I can't test it myself. Please let me know if the issue persists (or indeed, any other such Windows-specific bugs emerge!)

Thanks again and best wishes,
Peter

[bbosnja]

Hi Peter,

Thank you for quickly fixing this issue!! I have reinstalled the package and the problem was gone, so the fix worked :)

However, I have bumped immediately into the additional issue: during installation I got the message: "Skipping 6 packages not available: BiocParallel, UCell, CATALYST, Nebulosa, slingshot, ComplexHeatmap"

Hence, I have installed this packages separately.

Upon running MARMOT script, now I get the following errror:
"Error:
! ❌ Error: UCell is missing. Please check in the input settings chunk and make sure the variable exists/isn't commented out.

Quitting from MARMOT_Pipeline_LF-vs-HF.Rmd:96-116 [check_variables]", irrespectively if UCell package is loaded or not.

I am not sure whether this is also related due to running the package on a Windows machine.

Best,

Berislav

[bbosnja]

Hi Lydia and the whole MARMOT team, Thank you for your email. I have followed the instructions and could run the MARMOT with our data 😊 It is very nice package you have built, congrats! Shiny app gives a nice overview of the results and a lot of possibility to get nice figures. However, I still have one issue. It is about the number of clusters FlowSOM currently uses. If I got it correctly, the MARMOTs are testing 20, 40 or 60 clusters, which overclusters our data. Could it be possible to include the step where the user can specify the number of clusters for FlowSOM analysis? Best, Berislav Von: Lydia Kirsche ***@***.***> Gesendet: Dienstag, 14. April 2026 19:01 An: peterleary/MARMOT ***@***.***> Cc: Bosnjak, Berislav Dr. ***@***.***>; Author ***@***.***> Betreff: Re: [peterleary/MARMOT] Error with opening Metadata (Issue #4) [Das Bild wurde vom Absender entfernt.]lydiakirsche left a comment (peterleary/MARMOT#4)<#4 (comment)> Hi Berislav, Thanks for the feedback, that was really helpful. We’ve just released a new version of MARMOT on the main branch, which now also includes a GUI application. I’ve tested this version on Windows, and both the GUI and the code-based workflow should now run without the issues you encountered. Could I ask you to completely remove your current installation of MARMOT (e.g. using remove.packages("MARMOT")), restart R, and then reinstall it fresh via: pak::pkg_install("peterleary/MARMOT") Alternatively, you can also install and run it via the new GUI app—please have a look at the updated instructions <https://github.com/peterleary/MARMOT> . Also, please make sure to download the updated metadata, as the new version depends on it. Please let us know if the problem persists after reinstalling, or if anything else comes up. The new Wiki<https://github.com/peterleary/MARMOT/wiki> might also be helpful. Thanks for using MARMOT!! You might actually be the first user of this brand-new version 🙂 Best, Lydia — Reply to this email directly, view it on GitHub<#4 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ASHXLTHWLVBZBEFEIUOPG4L4VZVGPAVCNFSM6AAAAACXTUI77WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DENBVG4ZDQMZRG4>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[lydiakirsche]

Hi Berislav,
Thanks a lot for your kind feedback—we’re really glad to hear that MARMOT is running with your data and that you find the Shiny app useful 😊

Also, apologies for the confusion regarding the previous message. I had posted a reply on GitHub but then discovered an issue with the new GUI app on Windows, so I deletd the reply here. However, you still have received it via email notifications.
Since you’re running the pipeline via R, this likely didn’t affect you though.

Regarding your question on FlowSOM clustering: you can actually define the k-value for the metaclusters yourself in the metadata sheet. The values 20, 40, and 60 are just example settings we included by default. You can modify these in the metadata and specify the number of clusters you would like to use for your analysis. (e.g. just define 10, 18, 24 for kValuesIWant and then for knn e.g. 10). Also maybe try out a different Clustering method, I liked Phenograph/MPhenograph often better, especially when subgating to populations like Tregs etc. :)

Please let us know if anything is unclear or if you run into further issues!

Best,
Lydia

[bbosnja]

Hi Lydia,

This answers all my questions. I appreciate a lot your quick replies!

Best,
Berislav