two different junction type for the same circ?

[alexey0308]

Hi @tjakobi ,
is it correct to report two different junction types for the same circ coordinates?
The linear transcript counts are the same for both of them (see attached).
I think, only the first circ must be reported.

./CircCoordinates:3 63912588 63913225 ATXN7 1 + exon-exon
./CircCoordinates:3 63912588 63913225 N/A 2 - exon-exon

Regards,
Alex
bug.txt

[CDieterich]

Which data set are you using ?

[alexey0308]

It is from the MCF7 results.

grep 63912588 /prj/MCF7_RNA_turnover_Naarmann/preprocessing/workflow/DCC/CircCoordinates

[CDieterich]

Ok, then it is indeed stranded RNA-seq and a case that we should look at in detail.

[tjakobi]

I'm not sure that this is an DCC issue. Looking in the junctions files provided to DCC I can see that there is something on the '-' strand that barely gets over the threshold. If I remember the format of the junctions files correctly, the splice type reported there is also 2 which is passed through by DCC.