Why is there a big difference between kallisto and cellranger count result？

[aizhimin]

sample：
https://www.ncbi.nlm.nih.gov/sra/SRR8315735

kallisto result:

cellranger result:

[gaofan83]

The data you analyzed only has 25 million reads (people typically sequence 10X more reads). How many "cells" did you recover from the two pipelines?

[aizhimin]

@gaofan83 About 160 cells.

But why is there a big difference between kallisto and cellranger?

[gaofan83]

@aizhimin 160 cells from kallisto? How about cellranger?
My understanding is that the dataset you analyzed is a multi-omics data with both 5' expression and immune-seq. Are you sure you chose the correct 5' barcode whitelist as the input for kallisto?

[aizhimin]

@gaofan83 This is my cammand:
kb count -i /mnt/data/ref/transcriptome.idx -g /mnt/data/ref/transcripts_to_genes.txt -t 8 -m 16G --mm --h5ad --filter bustools --cellranger --gene-names --verbose --overwrite -x 10XV2 -o output/SRR8315735 SRR8315735_1.fastq.gz SRR8315735_2.fastq.gz

Where is the the correct 5' barcode whitelist ?

[gaofan83]

I am not familiar with kb. If you use kallisto/bustools, you will see there is an option for the whitelist. You can search 10X site and see whether 3' barcode whitelist (people typically call as 10xv2 or 10xv3) is the same as 5' barcode whitelist.
https://github.com/BUStools/bustools

[aizhimin]

@gaofan83 10xv2 Contains 5' barcode whitelist. Look at this image from 10x site。