autoCluster.batch {MetaCyto} | R Documentation |
A function that clusters the pre-processed fcs files from different studies in batch.
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)