After subsetting a seurat object number of variable features is wrong.

Question

[Assa Yeroslaviz]

When I subset a seurat object and start to reanalyze it, I can't change the number of variable features from the older (bigger object).

seuObj
An object of class Seurat 
26379 features across 56411 samples within 1 assay 
Active assay: RNA (26379 features, 2000 variable features)
 3 layers present: data, counts, scale.data
 3 dimensional reductions calculated: pca, umap, tsne

After subsetting and cleaning the object, I re-do also the normalization and finding the variable features, but it stays at 2000, even though I can see in the plot, that I have a lot more.

export <- subset(x = seuObj, idents = c(1,2,11) )
export <- NormalizeData(export, normalization.method  ="LogNormalize", scale.factor = 10000) 
export <- FindVariableFeatures(export, selection.method = "mean.var.plot", mean.cutoff = c(0.0125, 3), dispersion.cutoff = c(0.5, Inf))

export
An object of class Seurat 
26379 features across 13384 samples within 1 assay 
Active assay: RNA (26379 features, 2000 variable features)
 2 layers present: counts, data

How can I reset (change)this value?

[Bastien Hervé]

It might just be an issue of update in the printing of the class Seurat

2000 variable features

Double check the actual number of variable features you actually have in the object

length(VariableFeatures(export))

[Rogerio Ribeiro]

Did you find a workaround for this?