search

snijderlab / stitch

Template-based assembly of proteomics short reads for de novo antibody sequencing and repertoire profiling
MIT License
22 stars 3 forks source link

Germline sequence above the predicted sequence in the Alignment pile-up #160

Closed avilella closed 2 years ago

avilella commented 2 years ago

Hi,

In the same way that the report shows the germline and predicted proteins aligned to each other at the beginning of the report, it would be useful to show the Germline sequence above the predicted sequence further down in the Alignment pile-up, possibly colouring the peptides according to them mismatching the germline.

In this way, if there are lots of peptides mismatching the germline in a region, one would be confident that the difference against the germline is real, but if there is a combination of matches and mismatches in the different peptides in another region, one would be less confident of that difference.

I hope it makes sense

douweschulte commented 2 years ago

Just for clarity I will be a bit verbose. In any template or recombination detail page there is first the section with the aligned sequences (Annotated consensus sequence): image Further down (in the Alignment) we find the aligned peptides, here in dark blue the reference sequence is shown. In the case of a template matching step (any from the sections starting with Segment like Segment IGHV) the reference structure is the germline sequence. In the case of a recombined alignment (from the recombination table) is the consensus sequence from the previous template matching step. image

Providing the germline sequence besides the reference sequence in the case of recombined templates would be possible, but I fear it is confusing for users if this would be implemented without proper care. Because the alignment is made to the reference structure and not the germline.

The colouring for differences would be possible. I would propose a toggle between colouring according to difference with the reference and difference with the germline, if this is needed. With two different colours used for both. image

douweschulte commented 2 years ago

Closing the issue because the discussion has stalled and the proposal was not clear enough to drive implementation. If the discussion springs back to life the issue can be openend again.