Good news: no more arrow binary for cell ID! Conversion of those is really slow, such a pain in the ass. The images are now different, with a morphology_focus directory with multiple images, probably from different z-planes. Need a way to determine the version, since for Xenium Onboard Analysis v2 I would prefer to read the parquet file.
Also need to deal with multiple nuclei in one cell. I would need to wait for the release of v2 datasets, just to get the new label_id column which is required to make MULTIPOLYGON geometries from the nuclei.
See https://www.10xgenomics.com/datasets/preview-data-ffpe-human-lung-cancer-with-xenium-multimodal-cell-segmentation-1-standard
Good news: no more arrow binary for cell ID! Conversion of those is really slow, such a pain in the ass. The images are now different, with a morphology_focus directory with multiple images, probably from different z-planes. Need a way to determine the version, since for Xenium Onboard Analysis v2 I would prefer to read the parquet file.