Z-score normalization scales each feature (gene) across all cells to have a mean of 0 and a standard deviation of 1. This function allows for optional weighted univariance normalization, where cells are grouped by a categorical variable.
Value
A matrix of Z-score normalized gene expression data, with rows representing genes and columns representing cells.
Examples
library(Seurat)
library(data.table)
library(SpatialEcoTyper)
library(googledrive)
drive_deauth() # no Google sign-in is required
drive_download(as_id("1CoQmU3u8MoVC8RbLUvTDQmOuJJ703HHB"),
"HumanMelanomaPatient1_subset_counts.tsv.gz", overwrite = TRUE)
#> File downloaded:
#> • HumanMelanomaPatient1_subset_counts.tsv.gz
#> <id: 1CoQmU3u8MoVC8RbLUvTDQmOuJJ703HHB>
#> Saved locally as:
#> • HumanMelanomaPatient1_subset_counts.tsv.gz
scdata <- fread("HumanMelanomaPatient1_subset_counts.tsv.gz",
sep = "\t",header = TRUE, data.table = FALSE)
rownames(scdata) <- scdata[, 1]
scdata <- as.matrix(scdata[, -1])
tmpobj <- CreateSeuratObject(scdata) %>%
SCTransform(clip.range = c(-10, 10), verbose = FALSE)
#> Warning: Data is of class matrix. Coercing to dgCMatrix.
seurat_version = as.integer(gsub("\\..*", "", as.character(packageVersion("SeuratObject"))))
if(seurat_version<5){
normdata <- GetAssayData(tmpobj, "data")
}else{
normdata <- tmpobj[["SCT"]]$data
}
# Z-score normalization
znorm_data <- Znorm(normdata)
#> Centering and scaling data matrix
head(znorm_data[, 1:5])
#> HumanMelanomaPatient1__cell_3655 HumanMelanomaPatient1__cell_3657
#> PDK4 -0.3183628 2.5919742
#> TNFRSF17 -0.1060717 -0.1060717
#> ICAM3 -0.4436469 -0.4436469
#> FAP 3.2256642 -0.2634737
#> GZMB -0.2174644 -0.2174644
#> TSC2 -0.1316410 -0.1316410
#> HumanMelanomaPatient1__cell_3658 HumanMelanomaPatient1__cell_3660
#> PDK4 2.5919742 -0.3183628
#> TNFRSF17 -0.1060717 -0.1060717
#> ICAM3 -0.4436469 -0.4436469
#> FAP -0.2634737 -0.2634737
#> GZMB -0.2174644 -0.2174644
#> TSC2 -0.1316410 -0.1316410
#> HumanMelanomaPatient1__cell_3661
#> PDK4 -0.3183628
#> TNFRSF17 -0.1060717
#> ICAM3 -0.4436469
#> FAP -0.2634737
#> GZMB -0.2174644
#> TSC2 -0.1316410
# Weighted Z-score normalization
drive_download(as_id("12xcZNhpT-xbhcG8kX1QAdTeM9TKeFAUW"),
"HumanMelanomaPatient1_subset_scmeta.tsv",
overwrite = TRUE, verbose = FALSE)
#> Warning: The `verbose` argument of `drive_download()` is deprecated as of googledrive
#> 2.0.0.
#> ℹ Set `options(googledrive_quiet = TRUE)` to suppress all googledrive messages.
#> ℹ For finer control, use `local_drive_quiet()` or `with_drive_quiet()`.
#> ℹ googledrive's `verbose` argument will be removed in the future.
scmeta <- fread("HumanMelanomaPatient1_subset_scmeta.tsv",
sep = "\t",header = TRUE, data.table = FALSE)
wtdznorm_data <- Znorm(normdata, groups = scmeta$Region)
head(wtdznorm_data[, 1:5])
#> 6 x 5 Matrix of class "dgeMatrix"
#> HumanMelanomaPatient1__cell_3655 HumanMelanomaPatient1__cell_3657
#> PDK4 -0.27144312 2.29194978
#> TNFRSF17 -0.09363659 -0.09363659
#> ICAM3 -0.38535691 -0.38535691
#> FAP 2.81145245 -0.22714128
#> GZMB -0.19294377 -0.19294377
#> TSC2 -0.11586941 -0.11586941
#> HumanMelanomaPatient1__cell_3658 HumanMelanomaPatient1__cell_3660
#> PDK4 2.29194978 -0.27144312
#> TNFRSF17 -0.09363659 -0.09363659
#> ICAM3 -0.38535691 -0.38535691
#> FAP -0.22714128 -0.22714128
#> GZMB -0.19294377 -0.19294377
#> TSC2 -0.11586941 -0.11586941
#> HumanMelanomaPatient1__cell_3661
#> PDK4 -0.27144312
#> TNFRSF17 -0.09363659
#> ICAM3 -0.38535691
#> FAP -0.22714128
#> GZMB -0.19294377
#> TSC2 -0.11586941