search

PacificBiosciences / pbbioconda

PacBio Secondary Analysis Tools on Bioconda. Contains list of PacBio packages available via conda.
BSD 3-Clause Clear License
243 stars 44 forks source link

pigeon make-seurat : cell count is different from the result from isoseq correct #652

Closed Hyeyeonggg closed 4 months ago

Hyeyeonggg commented 5 months ago

My data is from single cell MAS-IsoSeq.

After cell calling with default knee method, the count of real cell was 1600. So I used the percentile method with 99 and got 3200 cells. But after pigeon make-seurat and making SeuratObject, I found the cell count 10700.

With the knee method, SeuratObject also has 1600 cells. With the percentile 98, SeuratObject has 11000 cells same with the output of isoseq correct.

I did the same command on above all examples.

I want to know why this discrepancy came from only in 99 percentile method.

Thank you.

Hyeyeonggg commented 5 months ago

For this issue, i tried to find where the difference generated.

I believe that isoseq groupdedup execution uses rc tag in BAM files from isoseq correct step.

I checked the cells having rc:i:1 which means predicted real cells in corrected.bam file generating from isoseq correct. --> 10747 But bcstats.tsv from isoseq bcstats with input file as corrected.bam showed the real cell count --> 3225

The other sample: 8497 vs 8686

Does bcstats.tsv filter more when deciding the real cells?

armintoepfer commented 4 months ago

Please reach out to support@pacb.com, as we can only help you with technical issues. Thank you.