I think there may be a bug here, which may be due to using participants instead of samples here - can you check out PT_9S6WMQ92 - this patient has two different events sequenced, one is stranded and one is polyA (medullo). Seems like we capture stranded in the plot but not the polyA. Caught my eye.
From this comment:
https://github.com/AlexsLemonade/OpenPBTA-analysis/pull/1698#discussion_r1142559984
For this issue, have a look at what's up!