Remove variable features in seurat object

[TiphaineCMartin]

Hi,

I want to rerun ScaleDate on all dataset.
I first select only the dataset of interest, but have still variable features from my previous object. I cannot remove selected variable features. I still have my scale.data on previous 3000 features instead of 35782.

Can you tell what I can do?

`

Create a new object without the scale.data layer

gcdata_Mac_diet_onlyMac_slim <- DietSeurat(
object = gcdata_Mac_diet_onlyMac,
assays = "RNA",
layers = c("counts", "data"), # Keep only 'counts' and 'data' layers, implicitly removing 'scale.data'
dimreducs = c("pca", "umap"), # Optionally remove all dimensional reductions
graphs = NULL # Optionally remove all graphs
)
gcdata_Mac_diet_onlyMac_slim
An object of class Seurat
35782 features across 26291 samples within 1 assay
Active assay: RNA (35782 features, 3000 variable features)
2 layers present: data, counts
2 dimensional reductions calculated: pca, umap
VariableFeatures(gcdata_Mac_diet_onlyMac_slim) <- NULL

features_null <- features[!features %in% features]
VariableFeatures(object = gcdata_Mac_diet_onlyMac_slim) <- features_null
gcdata_Mac_diet_onlyMac_slim
An object of class Seurat
35782 features across 26291 samples within 1 assay
Active assay: RNA (35782 features, 3000 variable features)
2 layers present: data, counts
2 dimensional reductions calculated: pca, umap

gcdata_Mac_diet_onlyMac_slim <- ScaleData(gcdata_Mac_diet_onlyMac_slim,

•                                       vars.to.regress = "percent.mt")
  

counts.matrix <- as.data.frame(t(as.matrix(gcdata_Mac_diet_onlyMac_slim@assays$RNA@layers$scale.data)))

dim(counts.matrix)
[1] 26291 3000`