Colors per channel are not the same than original .mrxs file

[ltalignani]

Dear all,

I am using bioformats2raw and raw2ometiff (latest versions installed with conda) to convert fluorescence image, coming from a 3Dhistech slides scanner, in .mrxs format.

bioformats2raw --progress lame1.mrxs zarr-pyramid-lame1/

followed by

raw2ometiff zarr-pyramid-lame1/ pyramid.lame1.ome.tiff

When I try to open the ome.tiff generated in Qpath, the channel colors are inverted

Image opened by QPath:

Original slide, in .mrxs format:

Is there a flag that can manage color per channel?

Thanks for your help

[melissalinkert]

Thanks for reporting this, @ltalignani. By latest versions, do you mean bioformats2raw 0.6.1 and raw2ometiff 0.4.1?

There isn't a flag to control which colors are assigned to which channels - that information is supposed to be automatically detected from the Slidedat.ini file in the .mrxs dataset. From the screenshots, it's difficult to tell if this is a problem with the channel colors being set incorrectly, or a problem of the channels themselves being out of order. We would need to see the original dataset that demonstrates this problem in order to investigate.

[ltalignani]

Thanks @melissalinkert for your answer.

I have the following information:
Bioformats2raw Version = 0.6.1
Bio-Formats version = 6.12.0
NGFF specification version = 0.4

Raw2ometiff Version = 0.4.1.

And you can download the .marxs files at this address:

https://filesender.renater.fr/?s=download&token=d4b7321c-eece-4f9e-939c-35e3e8f464d4

Sorry, the .dat files are not in a specific directory. You have to create it.

Thanks a lot

[melissalinkert]

And you can download the .marxs files at this address:

https://filesender.renater.fr/?s=download&token=d4b7321c-eece-4f9e-939c-35e3e8f464d4

Unfortunately, I see an error message when trying to access the data:

Could you please check the link, or upload to https://zenodo.org/ instead?

[melissalinkert]

As noted above, I strongly suspect there is a little more to the issue than just incorrect colors. We'll definitely need to see the data in order to investigate further, though.

In the absence of a working download link, closing for now. @ltalignani, if you're able to upload the dataset somewhere else, let us know the new link and this issue can be re-opened.