A possible explanation for issues with Clustree not finding enough clustering data

[beazors]

Hello,

Thanks for developing this package, I have used it for several years. Recently I went to test my old Clustree code to help a colleague with their data, and found that I had the same issue as others mentioned in issues #51 and #86.

I had previously written code to run FindClusters at different resolutions (it was written as a loop), and the resulting factor of cluster identities at each resolution used to be saved by default into the object's meta.data (e.g., "integrated_snn_res0.1").

In testing out my old code, I realized that this no longer happens. It looks like Seurat was modified in later updates such that, after clustering, the IDs and SNN are no longer stored, see here and here.

I am now trying to figure out how to write the clustering IDs to the metadata and having an unrelated problem with that (just getting the writing of the name in meta.data to work within my loop), but I wanted to point out that I think this is why people have had this issue. Clustering data at different resolutions is no longer automatically saved in meta.data.