How does Psiclass generate the psiclass_vote.gtf file?

[dgodin19]

Hi,

I was able to run psiclass to create gtfs for each of my samples, but was wondering how a consensus was calculated for the psiclass_vote.gtf file.

[mbeavitt]

From the paper: https://www.nature.com/articles/s41467-019-12990-0#Sec10

"Selecting a global set of transcripts.

PsiCLASS selects a set of meta-annotations from the individual samples’ sets of transcripts by voting. Each transcript is assigned a score equal to the transcript’s average estimated abundance across the samples (i.e., every sample has one vote, weighted by the transcript’s abundance level; default cutoff: 1.0). As different cutoff parameters values might work best for data with specific characteristics, the user can readily adjust or recalibrate the voting parameters postassembly, starting from the already computed sets of transcripts for the individual samples."

[mourisl]

@mbeavitt Thanks for answering. @dgodin19 To be more accurate, it calculates the average coverage value across the samples (the "cov" field in the GTF file), and you can use the "--vd" option to control the threshold.