Detecting duplicated UMI in BAM file

Question

[tien]

Dear all,

I'm trying to remove reads in BAM file resulted by Cell Ranger that have the same UMI per CB and keep only one, since I need to filter by some condition and count uniquely satisfied UMI. I'm using pysam and function is_duplicate() to detect those reads. I'm wondering if it is enough to detect duplicated reads? And if sam flag is updated after correcting UMI and cell barcode in cell ranger?

And also, if someone know how sam flag recognizes duplicated reads?

Thanks for your help.

[rpolicastro]

You'll need to check whether cell ranger adds the duplicate flag for that to work. If not you would need to filter by the cell barcode and UMI tags.

[tien]

Hi, thanks for your respond. Cell ranger has such flag (at least, there exists reads with samflag = 1024). But I'm not sure if it's enough? Because I observe something like in this post UMI and CB for the same sample but different lane by 10X single-cell RNA sequencing