search

cellgeni / sceasy

A package to help convert different single-cell data formats to each other
GNU General Public License v3.0
363 stars 53 forks source link

Transformation from AnnData/Loom produces SCE object that R doesn't seem to like #17

Closed pcm32 closed 3 years ago

pcm32 commented 3 years ago

Input dataset: https://humancellatlas.usegalaxy.eu/datasets/11ac94870d0bb33a4d823102dcb034bc/display?to_ext=loom

Result dataset was specifically https://humancellatlas.usegalaxy.eu/datasets/11ac94870d0bb33a005bfb8cba4fb8dd/display?to_ext=rdata (although it is an RDS, long story).

Entire history here: https://humancellatlas.usegalaxy.eu/u/pmoreno/h/atlas-scanpy-scmap-e-geod-94641-2

Will add more details on the R part.

a-solovyev12 commented 3 years ago

Parsing the object in R via readRDS() and then getting its attributes causes the following: 

> query_sce
class: SingleCellExperiment 
dim: 
metadata(5): neighbors pca raw.var raw.X louvain
Error in getClass(element.type) : “SimpleAssays” is not a defined class
nh3 commented 3 years ago

hmm, I am having the same error querying the content of your rds file. However, https://humancellatlas.usegalaxy.eu/datasets/11ac94870d0bb33a005bfb8cba4fb8dd/stdout seems to suggest the rds is normal under sceasy_convert's container environment. I also tried running the conversion code in my local environment using the input loom and I can read normally the content of the converted rds in the same environment. My guess is that the bioconductor/SingleCellExperiment versions are different/incompatible between sceasy_convert and scmap_preprocess_sce.