-
Notifications
You must be signed in to change notification settings - Fork 14
Description
Hi all,
I want to suggest some changes to accommodate the updated XOA output, since the XOA v4 was released this September.
- Fortunately, the naming of the morphology focus images are now named by channel AND marker, thus, the .get_xenium_images should be updated to be able to read the Xenium images processed with newer XOA versions. Maybe somthing like
} else if (major_version == 4L) { stains <- c("dapi", "atp1a1_cd45_e-cadherin", "18s", "alphasma_vimentin") img_fn <- paste0("ch000", 0:3, "_", stains, ".ome.tif") img_fn <- file.path(data_dir, "morphology_focus", img_fn) if_exists <- dir.exists(file.path(data_dir, "morphology_focus")) if (!if_exists) { warning("morphology_focus images not found") }
would do the trick? And maybe adjust the stop_message in the .get_imgData function? It could state the concatenated path that isn't valid. Would be helpful when trouble shooting.
2) Not sure about the problem yet and if it is really related with the XOA release, but creating/writing the segmentations to the _sf.parquet doesn't work. <\del>
>>> Cell segmentations are found in .parquet` file(s)
Reading cell and nucleus segmentations
Making MULTIPOLYGON nuclei geometries
Making POLYGON cell geometries
Sanity checks on cell segmentation polygons:
..found 221 cells with (nested) polygon lists
..applying filtering
Checking polygon validity
Saving geometries to parquet files
Error: IOError: Unexpected end of stream`
But I will check if there is some problem with my data or any kind of mismatching versions. If you have an idea, I'd appreciate a hint :)
<\del>
Thank you and have a nice weekend :)