Groups keyboard shortcuts have been updated
Dismiss
See shortcuts

MAJIQ-SPEL

71 views
Skip to first unread message

Noa Wigoda

unread,
Jun 30, 2024, 1:12:02 AM6/30/24
to Biociphers
Hi,
I used MAJIQ-SPEL to view a LSV and the numbers on the splice graph is the same for both conditions for a few LSVs I checked.
The psi plots are not the same. 
Can you please explain.
Thanks,
Noa

MAJIQ-SPEL (2).zip

San Jewell

unread,
Jul 2, 2024, 10:41:51 AM7/2/24
to Biociphers
Hi Noa,

Unfortunately I'm not able to load the html that you have sent successfully, it seems to not be working properly offline. Do you think you can provide screenshots of the error?

One thing that comes to mind is that the first splice graph is only one of possibly multiple experiments, whereas depending on which quantification mode you are using, PSI/dPSI may depend on multiple tissues / samples / experiments. It's hard to make an assertion without guessing.

Thanks,
-San

Noa Wigoda

unread,
Jul 10, 2024, 4:38:50 AM7/10/24
to Biociphers

Hi,

The splice graph is in attached splice_grph.jpg, note that the numbers are the same for both conditions.

The psi plots for this splice graph is attached PSI_plots.jpg.

It looks like a difference in the blue and purple splice junctions, but the numbers on the graph are the same.

 The data was taken from deltapsi output in Voila.

 Gene Name: Gabpb1; LSV ID: ENSMUSG00000027361:t:126658470-126658577

LSV Direction: Target; Strand: -; Chromosome: chr2; Genome: mm10

My question: why are the numbers the same in both conditions, although this LSV was found in the deltapsi analysis?

Thanks,

Noa

splice_graph.JPG
PSI_plots.JPG

San Jewell

unread,
Jul 10, 2024, 3:48:24 PM7/10/24
to Biociphers

Hi Noa,

I very much appreciate you following up with a clearer explanation of the issue. I've reproduced the problem with it and determined the issue to be only in deltapsi modes, and with voila rather than with majiq-spel. No data needs to be re-processed, only voila view will need to be re-run with a new version and the "copy lsv" button clicked again. 

I have released a new version of majiq/voila which fixes this issue. Please follow the same installation procedure to install / upgrade majiq/voila and let me know if you find the issue fixed. 

Thank you again for reporting it!

-San

Alessandra Longo

unread,
Jul 16, 2024, 4:28:05 AM7/16/24
to Biociphers
Hi,
I have the same problems and majiq and voila versions are  2.5.6.dev1+g8423f68b, i need to re-install them to fix the bug?
Many thanks in advance
Ale

San Jewell

unread,
Jul 23, 2024, 10:45:03 AM7/23/24
to Biociphers
Hi Ale,

Yes, I only released the hotfix recently and if your error is the same I would first try re-cloning / installing the new version. If the problem persists please post back with a screenshot of your error and I will check if perhaps there is a different case which is problematic.

Thanks
-San
Reply all
Reply to author
Forward
0 new messages