Open JungnamChoLab opened 3 months ago
Hi - thanks for reaching out.
Do you have any suggestions for how we can replicate this issue? Is there a publicly-available sample we can compare with?
Are you working with outputs from AA or from AmpliconClassifier (AC)? Keep in mind that AA's graph & cycles file are both 0-based coordinates and fully closed intervals. However, I don't see in the SVABA Readme if they are 0 or 1 based.
AC on the other hand reports using a 0-based, half-open counting system like UCSC.
Thanks, Jens
Hi Jens,
Thank you so much for giving me answers! I think you fully addressed my second question. I worked with the output from AC. For the first issue, I worked with a not yet public dataset. But you gave the idea that I should compare with a public dataset. I will try and see.
Bests and Thanks, Ling
Hi there,
I have examined the junction sites of ecDNA cycles using SVABA. However, I have encountered an issue where some junction sites cannot be reassembled by SVABA. Upon further analysis, I have noticed that for the junctions detected, many of them have a 1 base pair difference compared to the real genome coordinate. I am uncertain whether this discrepancy is due to the format of the bed files or some other reasons. Does anyone have any insights or ideas?
Bests, Ling