Closed YagmurSimsekk closed 1 year ago
It is okay for me to drop that phrase in README and remove the convertToSE function.
I will have a closer look next week but could we reiterate the reasoning for not supporting the TreeSE objects?
It could be beneficial to support a standard data container (as simulation function outputs), and all the necessary data elements are very easy to take out when needed, in particular:
x <- assay(tse, "counts")
df <- colData(tse)
The advantage is that the whole analysis ecosystem to manipulate and visualize microbiome (time series) data will be then directly accessible.
If the package is not supporting TreeSE, then the package name should be changed from miaSim to something else because it is not part of the collection any more. We could potentially add wrappers on top of this to other independent packages in such case.
Can we here go through the arguments for the different options, I guess the output options now is to provide 1) matrices; or 2) TreeSE objects. @gaoyu19920914 do you have a clear preference?
Done in the ongoing PRs.
Currently, none of the simulation functions output TSE or SE, they are either matrix or list objects. If that's not going to change, I suggest deleting "The package is part of miaverse, and is based on the (Tree)SummarizedExperiment data container." in README section.
Also, should we still keep convertToSE function? I don't see a reason to keep it.
I can make the mentioned changes once it is confirmed by others.