search

HelenaLC / CATALYST

Cytometry dATa anALYsis Tools
66 stars 31 forks source link

How to remove certain markers/channels after clustering with CATALYST::filterSCE? #365

Closed denvercal1234GitHub closed 1 year ago

denvercal1234GitHub commented 1 year ago

Hi there,

Thank you for the package.

I had performed clustering and dimensionality reduction on my sce object and now proceed with differential expression analysis.

Would you mind advising me how to delete certain channel/markers (i.e., rows of the rowData) from my sce object with CATALYST::filterSCE?

The reason is otherwise some rows within rowData(sce) are actually not true markers but rather just some metadata I added previously, and if I don't remove them, they will be considered as markers in the differential testing.

I tried, but not sure if there is other ways to do it.

F37_sce <- filterSCE(F37_sce, !(channel_name %in% Non_marker_labels_toRemove_vector))

Thank you.

HelenaLC commented 1 year ago

This issue isn't related to CATALYST, but has to do with basic SingleCellExperiment data handling. You can subset markers (rows) the same way as for any array-like object, e.g., data.frame or matrix, i.e., via sce[..., ] where ... is a logical, integer or character vector specifying which rows to keep. Please consult SingleCellExperiment and/or SummarizedExperiment's documentation to familiarize yourself with basic data handling for these classes.