DillonHammill / CytoExploreR

Interactive Cytometry Data Analysis
64 stars 13 forks source link

Cyto_transform is having problems with my gatingset #209

Open kim619 opened 5 months ago

kim619 commented 5 months ago

Hi I am trying to transform my data using the cyto_transform function but it encounters into error possibly due to the parameters used in cyto_transform (e.g, w, m)

Here are the codes


 gs<- cyto_transform(gs,channels=c("FITC-H","PE-H","FITC-A","PE-A","FSC-H","SSC-H","FSC-A","SSC-A","Time"),
                  display=10000,
                    trans=NULL,
                    type="log",
                    select="pCon08-G4.fcs",
                    m=5,
                    w=0.5,
                    widthBasis=80,
                     axes_limits="data")
Error: 'trans' must be a list of transformer objects (generated by scales::trans_new method)

I then tried using different parameters:

trans_logicle <- cyto_transformer_logicle(gs)
Error in .lgclTrans(x, p, ...) : w is negative!Try to increase 'm'

 trans_logicle <- cyto_transformer_logicle(gs,
                                          m=10,
                                           w=-1000)
Error: 'trans' must be a list of transformer objects (generated by scales::trans_new method)
 trans_logicle <- cyto_transformer_logicle(gs,
                                           m=10,
                                           w=1)
Error: 'trans' must be a list of transformer objects (generated by scales::trans_new method)
 trans_logicle <- cyto_transformer_logicle(gs,
                                           channels = "PE-A",
                                           m=20,
                                           w=10)
Error: 'trans' must be a list of transformer objects (generated by scales::trans_new method)

I also tried using biexponential transformation:

 # Default biexponentail transformer
 trans_biex <- cyto_transformer_biex(gs)
Error: 'trans' must be a list of transformer objects (generated by scales::trans_new method)

Same issues persist, please let me know what do the parameters mean since I don't know their meaning it would be hard to tweak around or what can I do about this, thanks a lot!

rwbaer commented 3 months ago

I am not an expert so take my comment with a grain of salt. When I have gotten this message, it often seems that I have set the gain on my flow channel a little too high so that my unstained cells are above the first decade.

Empirically, I have ended up setting m=6 or m=7 or going with an arcsinh transform in these circumstances to make the error message go away.

Here is where to read a little more on the theory of the transform: Cytometry A. 2012 Apr; 81(4): 273–277. Published online 2012 Mar 12. doi: [10.1002/cyto.a.22030]