from_... and to_SummarizedExperiment

[ap--]

Hello everyone 👋

would it be possible to add support for conversion from and to SummarizedExperiment directly?
One of our researchers asked for it as a convenience feature.

library(anndataR)
library(SummarizedExperiment)
 
nrows <- 200; ncols <- 6
counts <- matrix(runif(nrows * ncols, 1, 1e4), nrows)
rowData <- DataFrame(feature_id=sprintf("ID%03d", 1:200))
colData <- DataFrame(Treatment=rep(c("ChIP", "Input"), 3),
                     row.names=LETTERS[1:6])
rse <- SummarizedExperiment(assays=SimpleList(counts=counts),
                            rowData=rowData, colData=colData)
 
a <- from_SingleCellExperiment(rse)  # this currently checks for SingleCellExperiment

Thanks for the great work,
Cheers,
Andreas 😃

[rcannood]

#212 was just merged, which adds better support for SingleCellExperiment. Please review the updated documentation for more information.

Having a from_SummarizedExperiment would be nice to have. Do you have a suggested implementation, based on the current available code?

[SZhengP]

Hello,
It seems that two more changes are needed to allow from_SingleCellExperiment to handle SummarizedExperiment objects.
I tested the changes and it now works.
#223

Thank you so much.

[lazappi]

What would be the motivation for going to/from a SummarizedExperiment directly? A SingleCellExperiment is more complete and a better match to anndata so I'm wondering what the advantage would be over getting users to do the conversion themselves with as(...)?

[SZhengP]

Hello,
Yes. Using as(se, "SingleCellExperiment") is the strategy we are using now.
It is more of a philosophical thing. Since our data is bulk-RNAseq data, we are reluctant to use SingleCellExperiment class.