Missing AS types in Voila Modulize

30 views
Skip to first unread message

Yihao ZHANG

unread,
Jul 26, 2024, 6:59:34 AM7/26/24
to Biociphers
Hi, 

I read the documents of Voila Modulize. It said that there should be an additional column for exitron. But the output does not contain exitron.tsv. Neither did summary.tsv have this column. 

May I ask how to identify exitron especially using your tool?

Thank you very much!
Yihao
Message has been deleted

Yihao ZHANG

unread,
Jul 26, 2024, 8:03:04 AM7/26/24
to Biociphers
In tsv_writer.py, your comments said: 

# just in case exitrons are ever quantified, somehow, *try* to get
# the exitron's junction quantification (won't exist for now... which is desired)

I don't understand what it means. That seems to be related to the absence of exitron.tsv in the folder. Would you mind explaining a little about this issue? 

San Jewell

unread,
Jul 31, 2024, 12:35:54 PM7/31/24
to Biociphers
Hi Yihao,

I spoke a bit with Caleb Radens, one of the main authors behind the modulizer method, here is the reasoning behind the change in functionality for exitrons:

Exitrons are not currently supported because MAJIQ doesn't currently save coverage info at the exon level, and thus these types of junctions are not captured/quantified by our LSV model. However, we are aware that exitrons may be of interest to users, and we are weighing pros/cons of adding support for exitrons in a future update (quantifying exons would increase compute time and file sizes for e.g.)

It is possible to get exitrons quantification using MAJIQ even at the current version but this requires a two-steps process: First identify those exitrons junctions from junction spanning reads (e.g. STAR's junction reads file), then add those junctions into the annotation, then feed that costume annotation to MAJIQ. This would make MAJIQ treat the in between regions as retained introns events which will be then quantified and visualized accordingly. We are considering adding such an processing pipeline for interested users but it should be very doable for bioinformatics inclined users to implement themselves.

Again, we will consider adding exitrons in at a later time but as saving the exon coverage information would significantly increase processing time and file sizes we are weighing our options here.

I hope it helps answer your question. 
Thanks,
-San
Reply all
Reply to author
Forward
0 new messages