circBase: understanding how information are organized in the database

[Revo]

Hi All,

hope I posted in the right section.

circBase (Glažar et al., 2014)⁠ is a database of circular RNAs (circRNAs) that also provides code to identify known and novel circRNAs in sequencing data (Memczak et al., 2013)⁠.

I'm trying to better understand how information are organized in the tables provided in circBase. Parsing the BED table storing the annotation for all the Homo sapiens circRNAs (available here) I have found that there are 92,368 unique circRNAs IDs. However, checking at the FASTA file of putative human spliced circRNA sequences (available here) I have found 140,790 unique IDs. All the IDs from the BED file are present in the FASTA but there are 48,422 circRNAs that seems to be not reported in the BED file.

I tried to write to the maintainer (first author) but I haven't received a reply yet.

Is there anyone who has experience with this database that can help me understand how this difference can be explained? Am I missing some part of the documentation describing the filter used?

Thank you very much for any help.

Best,

Massimiliano