BoxPlotMetaClust: Statistical comparison of metacluster population between...
In gautierstoll/CytoSOM: Easy use of FlowSOM

Description Usage Arguments Value Examples

BoxPlotMetaClust is a function that allow the statistical analysis, either in percentage or with a normalization, of the impact of experimental condition on each metacluster population. Box plot of maker sizes, for the different meta-cluster with statical analysis, output in a PDF file.

BoxPlotMetaClust(
  TreeMetaCl,
  Title,
  treatmentTable,
  ControlTreatment,
  BottomMargin,
  yLab,
  Norm = FALSE,
  Robust = TRUE,
  ClustHeat = TRUE
)

`TreeMetaCl`	FlowSOM tree with meta-clusters, constructed by buildFSOMTree
`Title`	prefix of the pdf file name
`treatmentTable`	data frame containing a column 'files',a column 'Treatment' (case sensitive) and a column "NormalizationFactor" for the optional normalization of the data
`BottomMargin`	size of the boxplots bottom margins, in order to be readable
`Norm`	if true, sizes are normlized according to column 'NormalizationFactor' of the treatment table. Otherwise, percentage is used
`Robust`	if TRUE, robust statistics is applied (dunn), otherwise linear-model/Tukey
`ClustHeat`	if TRUE, heatmaps are clustered
`Controltreatment`	name of the control treatment used for statistics
`ylab`	name of y-axis of boxplots (eg 'population size')

list containing sizes, pairwise comparison p-values, controle vs treatment p-values

1	BoxplotMetaClust(TreeMetaCl=CytoTree, Title="Experiment1",treatmentTable="ExperimentTable1.csv",ControlTreatment="control",BottomMargin=3,yLab="CD45+",Norm=FALSE,Robust=TRUE,ClustHeat=TRUE)