Best Practice for Demultiplexing C57BL6/J Mixed snRNA-seq Samples Using CellSNP + Vireo?

[AxisJu]

Hi all,

I am working on Demultiplexing snRNA-seq data from mixed C57BL6/J mouse samples using CellSNP + Vireo. I have multiple snRNA-seq datasets, e.g.:

Sample01 (a mixture of 3 mice)
Sample02 (another mixture of 4 mice)
Since I cannot directly use the Mouse Genome Project genotype data for other mouse strains, I believe I should use model2 to generate VCF files.

My question is:
What is the best strategy for model2b and model1a usage in my case? Should I:

A: Run model2b separately for Sample01 and Sample02, then use the respective VCF files as input for model1a in each sample independently?
or
B: Provide a list of both Sample01 and Sample02 to model2b, generating a shared VCF file that is then used as input for model1a across all samples?
or
C: Use model2a instead of model2b for generating the VCF file?

I would appreciate any guidance on the best approach. Thank you!