unexpected unalign reads in capture sequencing about mixcr HOT 3 CLOSED

Also, I found there were significant differences at the final clones' ReadsCount between DNA and RNA-seq for another sample. And the RNA data had more assemble clones. Was this result normal? I used to think gDNA and RNA data both can apply to analyze IG/TR rearrangement, but the DNA result is unexpected, Can you give some advices?

Thank you.

from mixcr.

Hi,

The exome-seq preset processes two types of reads:

1. Those that fully cover the CDR3 region.
2. Those that partially cover the CDR3 region.

Any reads that don't overlap with the CDR3 region are discarded. This is because, without covering the CDR3, it's challenging to reliably assign alignments to a specific clone. However, the approach differs for SingleCell analysis. In that context, even reads that don't cover the CDR3 are retained. This is due to the understanding that such a read would have originated from one of the clones within a particular cell - a certainty we don't have in bulk repertoire analysis.

The reads you have shared seem to only span portions of the V gene (like the UTR, Leader, and Intron), making it ambiguous as to which clones they should be attributed to.

However, a yield of 24.31% is quite commendable, especially for targeted exome-seq.
Typically, RNA-seq provides better yields because multiple target RNA molecules exist per cell. So, it's not uncommon to see more favorable results there. Nevertheless, for in-depth repertoire analysis, targeted TCR/BCR sequencing is essential. If possible, using UMI barcoding can further enhance precision by allowing error correction.

from mixcr.

Thank you for the clear interpretation!

from mixcr.