Please add the option to use clustering in colData

[david-priest]

Please add the option to use a custom clustering stored in the column data when the input data is an SCE with cluster codes.

[SamGG]

Hello,
I think the best way is to use the package in a step-by-step way.
In the code of generateClusters.R, the few final lines of code show how to add any clustering result to the d_se object.
Best.

diffcyt/R/3_generateClusters.R

Lines 168 to 172 in 39348fd

	clus <- factor(clus, levels = seq_len(n_clus)) # includes levels for any missing/empty clusters
	# store cluster labels in row meta-data
	rowData(d_se)$cluster_id <- clus

[SamGG]

I think I mis-read your question. I thought you have carried out a table of counts per sample and per cluster.

If the input of the diffcyt wrapper/function is an SCE object, you can set the clustering_to_use argument to inform about the clustering column.

Unfortunately, I think it will currently fail as reported in Cell populations in topTable() wrong