autoCluster.batch: Cluster the preprocessed fcs files from different studies in...
In hzc363/MetaCyto: MetaCyto: A package for meta-analysis of cytometry data

Description Usage Arguments Value Examples

A function that clusters the pre-processed fcs files from different studies in batch.

1
2
3

autoCluster.batch(preprocessOutputFolder,
  excludeClusterParameters = c("TIME"), labelQuantile = 0.95,
  clusterFunction = flowSOM.MC, minPercent = 0.05, ...)

`preprocessOutputFolder`	Directory where the preprocessed results are stored. Should be the same with the outpath argument in preprocessing.batch function.
`excludeClusterParameters`	A vector specifying the name of markers not to be used for clustering and labeling. Typical example includes: Time, cell_length.
`labelQuantile`	A number between 0.5 and 1. Used to specify the minimum percent of cells in a cluster required to express higher or lower level of a marker than the cutoff value for labeling.
`clusterFunction`	The name of unsupervised clustering function the user wish to use for clustering the cells. The default is "flowSOM.MC". The first argument of the function must take a flow frame, the second argument of the function must take a vector of excludeClusterParameters. The function must return a list of clusters containing cell IDs. flowSOM.MC and flowHC are implemented in the package. For other methods, please make your own wrapper functions.
`minPercent`	A number between 0 and 0.5. Used to specify the minimum percent of cells in the positive and negative region after bisection. Keep it small to avoid bisecting uni-mode distributions.
`...`	Pass arguments to clusterFunction

A vector of labels identified in the cytometry data.

#get meta-data
fn=system.file("extdata","fcs_info.csv",package="MetaCyto")
fcs_info=read.csv(fn,stringsAsFactors=FALSE,check.names=FALSE)
fcs_info$fcs_files=system.file("extdata",fcs_info$fcs_files,
                               package="MetaCyto")
# Make sure the transformation parameter "b" and the "assay" argument
# are correct of FCM and CyTOF files
b=assay=rep(NA,nrow(fcs_info))
b[grepl("CyTOF",fcs_info$study_id)]=1/8
b[grepl("FCM",fcs_info$study_id)]=1/150
assay[grepl("CyTOF",fcs_info$study_id)]="CyTOF"
assay[grepl("FCM",fcs_info$study_id)]="FCM"
# preprocessing
preprocessing.batch(inputMeta=fcs_info,
                    assay=assay,
                    b=b,
                    outpath="Example_Result/preprocess_output",
                    excludeTransformParameters=c("FSC-A","FSC-W","FSC-H",
                    "Time","Cell_length"))
# Make sure marker names are consistant in different studies
files=list.files("Example_Result",pattern="processed_sample",
                 recursive=TRUE,full.names=TRUE)
nameUpdator("CD8B","CD8",files)
# find the clusters
excludeClusterParameters=c("FSC-A","FSC-W","FSC-H","SSC-A",
                           "SSC-W","SSC-H","Time",
                          "CELL_LENGTH","DEAD","DNA1","DNA2")
cluster_label=autoCluster.batch(
              preprocessOutputFolder="Example_Result/preprocess_output",
              excludeClusterParameters=excludeClusterParameters,
              labelQuantile=0.95,
              clusterFunction=flowHC)