Hello,

I've analysed my single cell RNAseq dataset in Seurat (v4). I am now wanting to run velocity analysis on it using CellRank (in python).

When running my analysis in Seurat I have renamed my clusters e.g :

p3.named <- RenameIdents(object = p3.named, `0` = "A", `1` = "B", `2` = "C",  `3` = "D" )

I then make the h5ad object as shown in the Seurat vignette using :

DefaultAssay(p3.named) <- "RNA"
SaveH5Seurat(p3.named, filename = "p3Namedvelocity.h5Seurat")
Convert("p3Namedvelocity.h5Seurat", dest = "h5ad")

Then, when I change onto python and import the object by doing the following and look at its obvs, the cluster names are the original (0, 1, 2 etc) :

adata = scv.read("p3Namedvelocity.h5ad")
adata

adata.obs["seurat_clusters"]
possorted_genome_bam_8Z2K8:AAACCCACAGACAAAT     1
possorted_genome_bam_8Z2K8:AAACCCAGTCTCGCGA     8
possorted_genome_bam_8Z2K8:AAACCCAGTTAGGGTG     0
possorted_genome_bam_8Z2K8:AAACCCATCTGTCTCG     1
possorted_genome_bam_8Z2K8:AAACGAAAGCTAGATA     3
                                               ..
possorted_genome_bam_8Z2K8:TTTGTTGCATCCGGTG    12
possorted_genome_bam_8Z2K8:TTTGTTGGTACAGTCT     2
possorted_genome_bam_8Z2K8:TTTGTTGCATGCAGGA    10
possorted_genome_bam_8Z2K8:TTTGTTGGTGCGAGTA    11
possorted_genome_bam_8Z2K8:TTTGTTGTCTTGGATG     0
Name: seurat_clusters, Length: 6892, dtype: int32

I've also tried renaming the 'seurat_clusters' in Seurat by doing :

p3.named[["seurat_clusters"]] <- Idents(p3.named)

but it results in exactly the same thing happening.

Can anyone help me solve this please? Thanks