Question About BFB Positive and Cyclic Class Amplicons in AmpliconClassifier

[DinDenkk]

Hi,

Thank you for developing AmpliconClassifier—it has been an invaluable resource for my research. I have a question regarding the interpretation of results in a specific case.

After running AmpliconClassifier, I observed an amplicon that was classified as BFB positive and had an decomposition class of cyclic. I am curious whether such an amplicon can be interpreted as ecDNA based on this classification alone.

I appreciate your clarification or any guidance on this matter and look forward to your insights.

Best regards,

[jluebeck]

Hi, thanks for this great question.

Oftentimes we see that AA decomposes BFB structures into cyclic genome paths. This does not change the fact that the amplification is a BFB, rather it indicates that from a bioinformatic perspective the genome reconstructions ended up being genome cycles. This of course also happens naturally with ecDNA, but for BFBs the genome cycles happen for a completely different reason. BFBs that get a bioinformatically cyclic decomposition seem to do so when there are lots of high copy-number inverted duplications and AA will extract a path that starts and ends at the same location, appearing like a genome cycle. Those cases (BFB-positive and decomposition class Cyclic) do not imply ecDNA.

BFB-negative and decomposition class Cyclic would on the other hand imply ecDNA.

Thanks,
Jens