SCIBER is a simple method that outputs the batch-effect corrected expression data in the original space/dimension. These expression data of individual genes can be directly used for all follow-up analyses. SCIBER has four steps; each step has a clear biological meaning, and the algorithms used for them are k-means clustering, t-test, Fisher’s exact test, and linear regression, respectively, all of which are easily comprehensible
You can install the development version of SCIBER with the following instructions:
# install.packages("devtools")
devtools::install_github("RavenGan/SCIBER")The following example uses the pre-processed Human dendritic cell dataset [1] to perform batch integration.
Please note that for each data frame in the object meta,
there should be two columns named cell_id and
cell_type. For instance, let meta_i be a data
frame under meta, and there should be two columns
meta_i$cell_id and meta_i$cell_type. If the
cell type information is not available, any values put in
meta_i$cell_type should work.
library(SCIBER)
rm(list = ls())
set.seed(7)
data(HumanDC)
exp <- HumanDC[["exp"]]
meta <- HumanDC[["metadata"]]
# Specify the proportion for each query batch to integrate batches.
omega <- c()
omega[[1]] <- 0.6
res <- SCIBER(input_batches = exp, ref_index = 1,
batches_meta_data = meta, omega = omega, n_core = 1)
#> [1] "The available number of cores is 10. SCIBER uses 1 to perform batch effect removal."