ValueError : No object to concatenate (Convert .mcd files to .ome.tiff files)

[Sudhirkumarchauhan]

Hi, I am trying to use the pipeline with example data provided. Everything works until step4 and analysis folder with required sub-folders are created. but then I am stuck at step 5 (Convert .mcd files to .ome.tiff files). it leads to error "ValueError : No object to concatenate". Please see the attachments. I am working on Window10.

Any help is very well appreciated!

Thanks, Sudhir

[nilseling]

Thanks for posting this issue! @jwindhager could you easily tell where this is coming from? Otherwise I'll dig into it.

[jwindhager]

@nilseling From a quick glance, it seems that acquisition_metadatas is empty, i.e. no mcd files have been found

[nilseling]

@Sudhirkumarchauhan did you download the example data before running the script and are they located in the folder that you specified?

[Sudhirkumarchauhan]

Yes, I did downloaded the data before running the script and they are located in the specified folder. After running step 1, I do not see any output, is that normal? then it runs step2-4 and stuck at mcd to ome tiff conversion. I activated the imcsegpipe environment already in conda.

Thanks, Sudhir

[nilseling]

I was able to reproduce the error on my side. The issue comes from this line. @jwindhager I have no clue why this regular expression doesn't work. It won't detect any of the ZIP files.

[nilseling]

Hi @Sudhirkumarchauhan

I have fixed the script. Please give it another try.

[Sudhirkumarchauhan]

Thank you very much! Now the pipeline works fine for me without any issue.

Sudhir

[Sudhirkumarchauhan]

Hi Again, The pipeline worked fine with example files, but not with my own data where I have some errors. Could it be something wrong in my panel file? I can also share .mcd file if that's helpful to troubleshoot.

Thanks in advance!

Sudhir

panel.csv

[nilseling]

Hmm this looks like some issue with extracting the channel label. Could you set full for ArAr80 to 0 and see if that fixes the issue?

[Sudhirkumarchauhan]

Thanks for the suggestion. I changed full for ArAr80 to 0, but still get the same error.

Sudhir

[jwindhager]

Most likely, the channel_label column is empty for some of the included channels, or the column only contains numeric values - could you please verify?

[Sudhirkumarchauhan]

Hi, I did check the panel file with the .txt file associated with mcd (both attached here). I think I have included all channel names in panel file as they are in mcd. I could send .mcd file on email (??) if they help in troubleshooting. Since I have no problem with example data, I'm scratching my head over what's wrong with my files!!

Thanks in advance!

Sudhir

220704_Irene_M12_G1_tumor_LEtest1_1.txt panel.csv

[nilseling]

Hi @Sudhirkumarchauhan I can't see the issue. So maybe it's worth if you send me (nils.eling@uzh.ch) one example MCD file.

[Sudhirkumarchauhan]

Thanks, I have just sent you the files over email. Sudhir

[nilseling]

So the issue is that the CSV files for the channel metadata in each ometiff folder contain some nan in the channel_label column. I'm not sure yet where this is coming from. @jwindhager would you have time to look into it?

[Sudhirkumarchauhan]

Hi @nilseling @jwindhager Thanks for the help so far. Did you manage to look at problem? Regards, Sudhir

[nilseling]

Hi @Sudhirkumarchauhan

we are still working on it and would need a bit more time, sorry for that. In the meantime you might want to check out steinbock for preprocessing/analysis.

[Sudhirkumarchauhan]

Thanks. Unfortunately I do not have opportunity to use docker on my work pc. Whenever you manage to troubleshoot the problem, just let me know. Hope it is fixed soon.

[jwindhager]

I cannot reproduce the error, everything seems to be working for me. Did you by any chance forget to adapt the file_regex wildcard to something that matches the names of your zip files (e.g. file_regex = "*.zip")?

EDIT: apologies, was looking at the wrong script - can reproduce the error now!

[jwindhager]

Ok, so the problem is indeed the existence of unlabeled channels (i.e., channels for which the user did not provide a label).

@nilseling please take a look at PR #113

[nilseling]

Hi @Sudhirkumarchauhan and @JaydenGittens,

we have now released a new version of the pipeline which should resolve the issues that you had.

@Sudhirkumarchauhan please double check the entries in your panel.csv file. We believe the entry Dy160 - TIGIT should be Gd160 - TIGIT.

Could you please confirm that this works for you and we can close the issue?

[JaydenGittens]

Hi Nils

This pipeline now works for me. Thank you ever so much.

Best Jayden Gittens

From: Nils Eling @.> Date: Wednesday, 8 March 2023 at 16:44 To: BodenmillerGroup/ImcSegmentationPipeline @.> Cc: Gittens, Jayden @.>, Mention @.> Subject: Re: [BodenmillerGroup/ImcSegmentationPipeline] ValueError : No object to concatenate (Convert .mcd files to .ome.tiff files) (Issue #110)

Hi @Sudhirkumarchauhanhttps://github.com/Sudhirkumarchauhan and @JaydenGittenshttps://github.com/JaydenGittens,

we have now released a new version of the pipeline which should resolve the issues that you had.

@Sudhirkumarchauhanhttps://github.com/Sudhirkumarchauhan please double check the entries in your panel.csv file. We believe the entry Dy160 - TIGIT should be Gd160 - TIGIT.

Could you please confirm that this works for you and we can close the issue?

— Reply to this email directly, view it on GitHubhttps://github.com/BodenmillerGroup/ImcSegmentationPipeline/issues/110#issuecomment-1460481387, or unsubscribehttps://github.com/notifications/unsubscribe-auth/A4PYOAUEN35RBDSM25N4Y7LW3CZNDANCNFSM6AAAAAAR4OEU2Y. You are receiving this because you were mentioned.Message ID: @.***>

[nilseling]

Closing due to inactivity - please reopen if this is still an issue.