search

hemberg-lab / SC3

A tool for the unsupervised clustering of cells from single cell RNA-Seq experiments
http://bioconductor.org/packages/SC3
GNU General Public License v3.0
119 stars 55 forks source link

Feature request: option to specify assay used #66

Closed dkoppstein closed 6 years ago

dkoppstein commented 6 years ago

Currently, SC3 defaults to using logcounts in a SingleCellExperiment. However, often you want to keep the raw logcounts data separate from batch-corrected or TMM-normalized data, i.e. using limma or ComBat and then assigning it to the norm_exprs value (or some other assay). It would be useful to have an additional option to specify the assay used from the SingleCellExperiment object, similar to scater::runPCA's exprs_values argument.

wikiselev commented 6 years ago

Hello, thanks for your request! The easiest would be to write whatever you would like to use for clustering into the logcounts slot. We used to have the exprs_values argument but stopped using it at some point as it was not easy to deal with all possible scenarios. Hope it helps.

alfonsosaera commented 6 years ago

I also think adding something similar to thescater::runPCA's exprs_values argument would allow for more flexibility when using the package. Thanks!

pati-ni commented 6 years ago

@wikiselev it would be possible to specify an assay during initialization. However if nothing is specified I think the current behavior is fine.

wikiselev commented 6 years ago

@pati-ni you can try, but I remember the way I wrote it originally does not allow it to be flexible for this parameter...

wikiselev commented 6 years ago

@pati-ni has also left, so I don't think this will be implemented. You are welcome to create a pull request, but if not, please use the logcounts slot for what you want to achieve.