Annotation in develop branch produces different results

[tjakobi]

While the positions of detected CircRNAs are the same between master and develop branch, the annotation itself differs.

Example:

631069 631291 RP5-857K21.6 0 + intron-exon exon,gene,transcript 1 631069 631369 RP5-857K21.6 0 + intron-exon exon,gene,transcript

1 631069 631291 RP5-857K21.6 0 + intron-exon transcript,gene,exon 1 631069 631369 RP5-857K21.6 0 + intron-exon transcript,gene,exon
Additionally to the cases with different annotation order in the type2 column, there are others cases:

1 10095461 10105744 UBE4B 2 + exon-exon CDS,exon,gene,transcript 1 10117459 10119613 UBE4B 2 + exon-exon CDS,exon,gene,transcript

vs.

1 10095461 10105744 UBE4B 2 + exon-intron gene,transcript 1 10117459 10119613 UBE4B 2 + exon-intron gene,transcript

And also cases where the type is different:

1 24514314 24514567 RCAN3 2 + exon-exon CDS,UTR,exon,gene,start_codon,transcript 1 24514314 24531391 RCAN3 2 + exon-exon CDS,UTR,exon,gene,start_codon,transcript

vs.

1 24514314 24514567 RCAN3 2 + exon-intron gene,transcript 1 24514314 24531391 RCAN3 2 + exon-intron gene,transcript

Type exon-exon vs. exon-intron.