The raw RNA-seq data are downloaded from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE142201 .
The raw files are processed with nf-core/rnaseq and merged into a single summerized experiment
Then to reduce the file size only highly variable genes and very low expressing
genes are kept which is enough for the vignettes.

```r
# To reduce the file size

library(broadSeq)
# se <- readRDS(system.file("extdata","rat_vole_mouseSE_salmon.rds", package = "broadSeq"))
se <- readRDS("inst/extdata/rat_vole_mouseSE_salmon.rds")

test_se <- broadSeq::transformDESeq2(se,method = "vst"  )
varGene <- broadSeq:::getSelectedGene(genes = NULL,se = test_se,ntop = 5000, scaledAssay = "vst")
varGene <- rownames(se)[varGene]

keep <- (assays(se)[["counts"]] >= 3) %>% rowSums() >= 5
lowGenes <- rownames(se)[!keep]

knownGenes <- as.data.frame(rowData(se)) %>%
    dplyr::filter(!(Class %in% c("Other","Dev. process"))) %>%
    pull(symbol)

test_se <- se[unique(c(knownGenes, lowGenes, varGene))]
test_se <- broadSeq::transformDESeq2(test_se, method = "vst"  )
computedPCA_vst <- broadSeq::prcompTidy(test_se, scaledAssay = "vst", ntop = 500)

plotAnyPC(computedPCA = computedPCA_vst,  x = 1, y = 2,
          color = "species", shape = "stage", legend = "bottom")

plotAnyPC(computedPCA = computedPCA_vst,  x = 3, y = 2,
          color = "species", shape = "stage", legend = "bottom")


for (an in assayNames(test_se)) {
    mode(assays(test_se)[[an]] ) <- "integer"
}

saveRDS(test_se,"inst/extdata/rat_vole_mouseSE_salmon.rds")
```
