Another possibility would be to look at ratio between linear RNA reads vs circular RNA reads, and pick circRNAs that has the ratio above certain threshold. I recently (researching for the review) noticed that there is CircExplorer3-CLEAR pipeline from the same group that developed circExplorer2. Version 3 uses version 2 for circRNA detection but take linear RNA reads simultaneously to calculate above-mentioned ratios, if I understand it correctly. I don’t know whether this can be integrated to your pipeline and indeed useful, but I feel this is promising to characterize circRNAs for their relative importance, because the ratio (linear vs circular) seems to be highly regulated in various circRNAs including viruses’ circRNAs.
kopardev
commented
3 years ago
On 12/02/20 Takanobu wrote:
Hence, adding CLEAR will be a good idea.