Corner case bug in sequenza.results?

38 views

Skip to first unread message

Michael Knudsen

unread,

Sep 26, 2024, 8:55:18 AM9/26/24

to Sequenza User Group

Hi,

I am running Sequenza 3.0.0 in a production workflow. I have run it on many samples – hundreds – and the setup appears is really robust. However, I have just stumbled upon a sample causing this error in sequenza.results:

Error in data.frame(..., check.names = FALSE) :
arguments imply differing number of rows: 6151, 6150
Calls: sequenza.results -> cbind -> cbind -> data.frame
Execution halted

It appears after these three output files have been generated:

gc_plots.pdf
CP_contours.pdf
chromosome_depths.pdf

Maybe a corner case bug? I would be happy to assist with debugging information. I cannot share data, though.

Thanks,

Michael

Michael Knudsen

unread,

Nov 28, 2024, 10:16:15 AM11/28/24

to Sequenza User Group

Hi,

Quick update on this issue. I have managed to get around it by adding

seg.tab <- seg.tab[seg.tab$N.BAF > 1,]

after seg.tab is first defined in sequenza.results. In the rare cases where N.BAF is 1, the corresponding rows seem to be ignored by the baf.bayes function. I'm far from fluent in R, and I've not been able to figure out exactly why this is so.