I have used the normalizeBatch() function of the cydar package to apply batch correction to a mass cytometry dataset, based on an identical technical replicate present within each batch.

The cydar package manual describes how to convert the output of normalizeBatch() to a CyData object (by using unlist() followed by the prepareCellData() function). However, I wish to keep the data in a flowSet (or ncdfFlowSet) structure for use in a downstream pipeline based on the CATALYST package.

Could anyone tell me please if there is a method to achieve this, i.e. is it possible to do any of the following? 1. apply the output from normalizeBatch() back to the ncdfFlowSet files used in the function to normalise them whilst maintaining their ncdfFlowSet structure 2. convert the output from normalizeBatch() back to an ncdfFlowSet structure

Apologies if I am asking the impossible here, any help would be appreciated.

I would imagine that you could take each matrix in the unlisted output of normalizeBatch() and use it to construct a flowFrame. The resulting list of flowFrames could then be converted into flowSet. This may or may not be coercible into a ncdfFlowSet - I can't remember - but it shouldn't matter as a flowSet should work just as well in any downstream application.

library(cydar)
example(normalizeBatch) # just to get some stuff.

library(flowCore)
blah <- unlist(corrected, recursive=FALSE)
blah <- lapply(blah, flowFrame)

fs <- do.call(flowSet, blah)

Hopefully the downstream pipeline don't use any of the special parameter fields in the flowFrames.