JinmiaoChenLab / cytofkit

cytofkit: an integrated flow/mass cytometry data analysis pipeline
http://jinmiaochenlab.github.io/cytofkit/
57 stars 25 forks source link

Total # of clusters in the output FCS is different from the total in other output files #48

Open msenosain opened 6 years ago

msenosain commented 6 years ago

I was analyzing my data from the cytofkit_analyzedFCS file and when I uploaded the cytofkit.RData file into the shinnyApp I noticed that for each clustering algorithm that I run it appears to be one extra cluster than the number of clusters I get in the FCS file. For example: If I upload the FCS in to R and work with the exprs object as a data frame and I do the following:

sort(unique(dt_df$Rphenograph_clusterIDs))
 [1]  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 

It means that there are 16 Phenograph clusters. But then when I load the cytofkit.RData file into the shinnyApp, for this example, I would obtain 17 Phenograph clusters, and the tSNE plot contains 17 clusters, so any analysis that I do to the FCS file is useless if I want to use the tSNE plot to illustrate my findings. The total number of events in the FCS file is the input that I gave it (e.g. 10K cells) so there is no lost information, its just a different number of clusters (-1 exactly). Any solution?

Thanks