Looking for RNA-seq aligner that can identify not properly mapped paired

Question

[romain.lannes]

Hi

I would like to identify pairs in RNA seq that dot not map properly. In particular, I am interested on this kind of pattern:

--->---------- R1

---------->--- R2

instead of 
--->---------- R1

----------<--- R2

which I believe i can identify in the sam using the flag: read reverse strand or mate reverse strand

After running STAR I have a staggering 100% properly paired flag. I wonder if STAR per default only report properly paired read. If there is a way using STAR or any other tools to achieve this?

I believe in desperation I have the option to align R1 and R2 individually, but that incurs a lot of post precessing. And I would rather not reinvent the wheels.

I greatly appreciate your help, Romain

Answer 1

[Matthias Zepper]

There are definitively tools out there to call structural variants, including inversions. I am not sure if RNA-seq data is the appropriate input, since most tools are likely written for WGS data, but InvBFM could work nonetheless ?!? Also see this example to extract discordant pairs.