issue in Raw count processing 'Seurat_SCT_transform' function

[mbsai]

I tried to replicate the process for Pan et al study. However facing trouble while running the step 2
seurat_obj <- Seurat_SCT_process(seurat_obj, seurat_filter = FALSE,
sample_id = "pan_rpe004",
study_name = "pan_et_al")
I got the following error:
Warning: The following features are not present in the object: CDCA7, DTL, RAD51, RRM2, CDC45, CDC6, EXO1, E2F8, not searching for symbol synonyms
Warning: The following features are not present in the object: UBE2C, BIRC5, MKI67, PIMREG, CCNB2, CKAP2L, AURKB, BUB1, HJURP, CDCA3, CDC20, TTK, CDC25C, DLGAP5, CDCA2, CDCA8, HMMR, NEK2, CENPA, not searching for symbol synonyms
Error in h(simpleError(msg, call)) :
error in evaluating the argument 'x' in selecting a method for function 'rowMeans': subscript out of bounds
In addition: Warning message:
Layer ‘data’ is empty

When I looked deeper into the function,
I was getting this error while running the following line of the function
seurat_obj = CellCycleScoring(pan_p1_seurat_sct, s.features = s.genes,
g2m.features = g2m.genes)

[joseverdezoto]

Hey,

Sorry for the late reply! I think that might be a bug intrinsic to the transition to Seurat v5. I fixed that bug on the dev branch of our R package https://github.com/MillerLab-CPHG/scRNAutils/tree/dev.
You can install it using:
devtools::install_github("MillerLab-CPHG/scRNAutils", ref = "dev")

You can run the entire processing pipeline or just the SCTtransform function. Hope this helps but let me know if you have any issues.