Repository navigation
Empty output files - No event detected #587
Description
Hello,
I've tried to run DROP pipelines Aberrant Expression and Aberrant Splicing on my dataset of 62 samples generated in the same conditions from the same tissue. However, nothing have been detected for aberrant expression, and only 18 events have been detected for aberrant splicing. It sounds very low to me, nevermore that I have at least 30 samples among this ones with known splicing events. However, quality control seems to be ok.
Here are the config file and sample file that I use for pipeline.
config.txt
test3_drop.txt
Have you ever experiment this issue ?
Thank you so much for any help,
Laëtitia RIALLAND
Activity
vyepez88 commentedon Jan 22, 2025
Hi Laëtitia, with sample sizes < 100 we have observed many genes that do not pass the stringent FDR cut-offs. You can try modifying those cut-offs in the aberrantExpression and aberrantSplicing dictionaries or checking the full results tables of each module.
laetitiarialland commentedon Feb 17, 2025
Hi Vicente,
Thank you for your help. Actually, when I try to modify FDR cut-off especially with very permissive modifications (p-value cut off < 0.3), I still have no results, or only 1 / 2.
I've omitted to precise in my first message that my samples comes from targeted RNA seq, could it be a reason explaining this problem ?
Thank you for your help,
Kind regards,
Laëtitia
vyepez88 commentedon Feb 17, 2025
Hi Laëtitia,
do you mean nominal p-value or FDR? If it's FDR, a cut-off of 0.3 can still be harsh.
Try loading the
results_gene_all.tsvfile in the aberrant splicing results folder and check the nominal p-value and FDR of your junctions of interest.I don't have much experience with targeted RNA seq, but I doubt that matters.