gpertea/stringtie

Assembled transcripts lack canonical transcript, clearly visible in the alignment

Opened this issue · 1 comments

Hi there,

thank you for the amazing tool. While assembling transcripts from ONT data, I noticed the canonical transcript of a gene of interest was missing and seems to be cut in two at a specific junction. The alignment shows clear reads containing both sides of the junction.

I have tried various values for parameters

-f
-t
-R
-g
-a

but no luck in getting a transcript resembling the canonical one.

I used StringTie 2.2.1, command stringtie $f -o $dir/${x##*/} -p 6 -v -L. I would like to avoid guided assembly as my samples contain splice effects, which are often ignored when running with -G. I've included two IGV screenshots to clarify the issue.
overview
zoom

Any ideas what might be the issue?

Regards
Mattias

Adding to that.

I ran StringTie in short-read mode (not using params -R or -L), with the following command
stringtie $f -o $dir/${x##*/} -f 0 -t
which yielded the output in this IGV screenshot.
fix

Maybe this helps to identify what might cause the issue reported above. If you would like any further data or info, happy to provide.

Regards
Mattias