Wrapper function around CytoNorm.train
from the CytoNorm package.
Usage
train_cytonorm(
fcd,
batch_var,
remove_param = NULL,
seed = 91,
files = NULL,
data_path = NULL,
FlowSOM_param = list(nCells = 5000, xdim = 5, ydim = 5, nClus = 10, scale = FALSE)
)
Arguments
- fcd
flow cytometry dataset
- batch_var
Column name of batch variable from
fcd$anno$cell_anno
.- remove_param
Parameters/markers which should be excluded for learning the batch effect and training the model.
- seed
A seed is set for reproducibility.
- files
Vector of FCS file names of reference samples which are used for training the model. If
files == NULL
, all files contained in the fcd are used.- data_path
File path to folder where .fcs files contained in the fcd are stored. This parameter does not need to be provided, unless the folder where the .fcs files are stored has changed.
- FlowSOM_param
A list of parameters to pass to the FlowSOM algorithm. Default=
list(nCells = 5000, xdim = 5, ydim = 5, nClus = 10, scale= FALSE)
Details
train_cytonorm
train_cytonorm' takes a fcd as an input and learns the batch effect of a given batch variable across reference samples provided by the user using the CytoNorm algorithm. This function returns a fcd with the trained model which can be used as input for the run_cytonorm
function to normalize samples with the trained model.
See [Van Gassen et al., 2019](https://doi.org/10.1002/cyto.a.23904) for more details on CytoNorm.