Reads per gene vs annotation file generated with CIRCexplorer2

[andre-gabriel-42]

For instance, let's say that my reads per gene contains the following information for these two transcripts (Reads per gene file obtained with STAR):

**ENST00000361204**	0	0	**0**
**ENST00000255784**	842	0	**842**

So 0 reads per ENST00000361204 transcript and 842 reads per ENST00000255784 transcript.

And the file generated during annotation step contains lines like the following:

chr22	42276719	42290941	circular_RNA/19	0	+	42276719	42276719	0	0	0	4	277	170	169	118	0	4126	12401	14104	19	circRNA	SREBF2	**ENST00000361204**	10	11	12	13	chr22:42274127-42276719|chr22:42290941-42293055

chr22	42204878	42206295	circular_RNA/260	0	+	42204878	42204878	0	0	0	3	119	122	85	0	1004	1332	260	circRNA	CCDC134	**ENST00000255784**	2	3	4	chr22:42196770-42204878|chr22:42206295-42209267						

chr22	42204878	42209826	circular_RNA/81	0	+	42204878	42204878	0	0	0	5	119	122	85	182	72	0	1004	1332	4389	4876	81	circRNA	CCDC134	**ENST00000255784**	2	3	4	5	6	chr22:42196770-42204878|chr22:42209826-42221701

Questions for you:

1. It seems that a circRNA had 0 reads yet it was annotated (ENST00000361204). Why? What does it mean?
2. For ENST00000255784 transcript, it is annotated twice as two different circRNAs. How would I know the weight of this transcript? How sure am I that it would not be a 3rd transcript? Should I use bedfiles? At which step?
3. After the annotation step, how do you actually cross the information between the circRNAs that were aligned and annotated? How do you extract the exact coordinates from the reads per gene file?

Thanks a lot for your help.