Error when compensating

[marioleivahidalgo]

Hi there,

It is Mario. I am doing a cytometry data analysis. But, when I do the compensation, this message appears in R:

spillover(fcsfile)
Error in .local(x, ...) : No spillover matrix stored in that sample

What could be the possible causes? I think it could be related to the type of file? FCS2 or FCS3?

[mikejiang]

flowCore looks up these fcs keywords c("SPILL", "spillover", "$SPILLOVER") for spill over matrix, if none is present in fcs file, you get this error message, so it has nothing to do with fcs file format version

[marioleivahidalgo]

First time trying to analize in R.
I try with other archives, from an exercice, and it works.
I receive my FCS files from an exportation of FACSDIVA software. So, should I give some special request and considerations about the exportation of my files ?

[mikejiang]

ask them to include spillover matrix in one of these keywords according to the standard https://flowcyt.sourceforge.net/fcs/fcs32.pdf

[djhammill]

@marioleivahidalgo, in the event that the spillover matrix may have been attached to a different keyword, you can search the FCS file to locate keywords containing matrices as below. If a spillover matrix is found you can extract it and pass it to compensate() directly, otherwise as @mikejiang suggested you need to ensure that the spillover matrices are appropriately attached on the instrument.

# path to first FCS file in folder called "folder"
file <- list.files("folder", full.names = TRUE, pattern = ".fcs")[1]
# read into a flowFrame
fr <- read.FCS(file)
# extract list of keywords
kw <- keyword(fr)
# restrict to keywords containing a matrix
kw[sapply(ks, "is.matrix")]

If the output of the last command contains the spillover matrix you can inspect the keyword name (e.g. "sp"), extract it from the flowFrame and pass it to compensate() manually:

# spillover has been erroneously stored under keyword called "sp"
spill <- keyword(fr, "sp")
fr_comp <- compensate(fr, spill)

[SamGG]

IIRC there is no compensation matrix in FCS 2.0

[marioleivahidalgo]

Thanks, Samuel. I guess that. I tried with @djhammill recommendations but it seems there is no compensation matrix in my FCS files. By the way, there is a big difference in file size when I compare my files with others that I downloaded to make an exercice with this package. Mine are for far lighter than those of the exercice. I will ask what type of FCS are exporting on the FACS facility.