psi results

[Swethaa NG]

Hi,

Hope you are doing well! :)

I have two doubts regarding psi quantified results please. 

1. One is about junction coordinates. Attached HTML file shows exon skipping event for RBM3 gene. When I checked the coordinates, the green LSV showed chrX:48575667-48576314 and these account for exon2-3 and ex3-4 junctions for transcripts ENST00000376755.1 and ENST00000376759.8 respectively.

My doubt is that after I checked the transcript info, it does not look like exon skipping event. Please correct me if I understood incorrectly. The LSV seems to be undergoing canonical splicing? If yes, then how do I handle these type of results without having to manually check them? 

Also, is it possible to get the transcript information from psi quantification itself?

I think it is possible to see the different transcript IDs when deltapsi and/or heterogen quantified data are viewed in voila?

2. I have attached the tsv file for this sample as well. I want to compare the PSI values for  junctions that are commonly detected in a bunch of samples. 

2. a. To do so, I just have to compare the mean_PSI_per_lsv_junction (for the matching gene ids and lsv ids only) right? Should I consider the stdev as well?

2. b. Sorry, I know these are quite naive doubts. May I please know how else can one utilise the .tsv info? 

2.c. If I were to understand the different lsv types  (denoting possibly different splicing events for the same or many genes) for each sample, I should run voila modulize on each of their psi.tsv and get info from individual modulise run .tsv files, right?

Here are my commands: 

majiq build /proj/env2/gencode.v39.chr_patch_hapl_scaff.annotation.gff3 -c config.ini -o /proj/env2/polyA_M/SJdir/denovo_disable --disable-denovo --disable-denovo-ir --simplify 0.01 -j 9

majiq psi -o /proj/env2/polyA_M/Voila_psi -n 4974 4974.majiq --minreads 30

voila view splicegraph.sql /proj/env2/polyA_M/Voila_psi/4974.psi.voila

thank you!

Kind regards,

Swethaa