CCA integration

Question

[kayah]

I’m currently analyzing my data using the following code:

WAT <- readRDS("~/Desktop/GSE137869/GSE137869(AtoZ)/WAT.RDS")
WAT<- NormalizeData(object = WAT)
WAT<- FindVariableFeatures(object = WAT)
WAT<- ScaleData(object = WAT)
WAT<- RunPCA(object = WAT)
ElbowPlot(WAT)
WAT<- FindNeighbors(object = WAT, dims = 1:10)
WAT<- FindClusters(object = WAT)
WAT<- RunUMAP(object = WAT, dims = 1:10)
library(gridExtra)
p1<-DimPlot(WAT, reduction = 'umap' , group.by = 'orig.ident')
p2<-DimPlot(WAT, reduction = 'umap' , group.by = 'type', cols = c('red','blue'))
grid.arrange(p1, p2, ncol = 2, nrow = 2)
#####################\\\\
WAT.list <- SplitObject(WAT, split.by = 'sex')
View([e-mail])
for(i in 1:length(WAT.list)){
  WAT.list[[i]] <- NormalizeData(object = WAT.list[[i]])
  WAT.list[[i]] <- FindVariableFeatures(object = WAT.list[[i]])
}

features <- SelectIntegrationFeatures(object.list = WAT.list)
WAT.list <- lapply(WAT.list, function(x) {
  x <- ScaleData(x, features = features)
  x <- RunPCA(x, features = features)
  return(x)
})
anchors <- FindIntegrationAnchors(object.list = WAT.list, anchor.features = features)
WAT.integrated <- IntegrateData(anchorset = anchors)
WAT.integrated <- ScaleData(WAT.integrated)
WAT.integrated <- RunPCA(WAT.integrated)
WAT.integrated <- RunUMAP(WAT.integrated, dims = 1:30)
WAT.integrated <- FindNeighbors(WAT.integrated, dims = 1:30)
WAT.integrated <- FindClusters(WAT.integrated, resolution = 0.5)
DimPlot(WAT.integrated, reduction = 'umap', group.by = 'sex', label = TRUE, pt.size = 0.5)

for(i in 1:length(WAT.list)){
  WAT.list[[i]] <- NormalizeData(object = WAT.list[[i]])
  WAT.list[[i]] <- FindVariableFeatures(object = WAT.list[[i]])
}

However, when I run this code, I get the following error:

Error in [<-.data.frame(*tmp*, , i, value = c(0.00151975683890578, : replacement has 15601 rows, data has 16176)

Is there a good way to match the number of rows to resolve this issue? Also, the datasets for males and females have different sizes—do I need to perform any resizing to match them?

[bk11]

Is there any reason for trying to use old method in integrating your data? I would you recommend to use the new approach from Seurat V5 for this purpose. Please check the link below:

https://satijalab.org/seurat/articles/seurat5_integration

[kayah]

Thank you so much!