I want to predict circular RNAs - miRNAs interactions. I am interested in collecting/extracting this information from available resources and/or performing co-expression network analyses.

The issues/challenges that I am facing are (1) Universal annotation of circular RNA? I don't know if it exists; if yes, does it be consistent in various databases, (2) What are the latest databases that store interaction either predicted or experimentally validated? (3) for co-expression analyses of miRNA-circRNA, what normalization count should I be using? CPM (Count per million) or others (l learned that length based normalization methods are not appropriate for miRNA)? And normalization count of back splice junction reads (for circRNA)?

One relevant question is posted here, which is more specific to the use of cancer-specific cirRNAs-miRNAs.

Also looking for relevant articles to these analyses.