Question About Filtering Differential Alternative 3' Splice Site Usage in MAJIQ V3

6 views
Skip to first unread message

X L

unread,
Jul 22, 2026, 9:23:27 AMJul 22
to Biociphers

Dear All,

I am using MAJIQ V3 to analyze differential alternative splicing between two groups of short-read RNA-seq data (control vs treatment), and I am particularly interested in differential alternative 3' splice site usage between the two groups. After running MAJIQ, I obtained a results.deltapsi table with the following columns: seqid, strand, gene_name, gene_id, event_type, ref_exon_start, ref_exon_end, start, end, is_intron, other_exon_start, other_exon_end, is_denovo, ref_exon_denovo, other_exon_denovo, event_denovo, dpsi_mean, dpsi_std, probability_changing, probability_nonchanging, control_1_raw_psi_mean, control_1_raw_psi_std, control_1_bootstrap_psi_std, control_1_raw_coverage, treat_1_raw_psi_mean, treat_1_raw_psi_std, treat_1_bootstrap_psi_std, and treat_1_raw_coverage.

I am new to MAJIQ and understand that it was designed to detect LSVs rather than specific event types. Is there a way to filter for differential alternative 3' splice site usage from the results.deltapsi table, or from other MAJIQ output tables?

Thanks,
Xiao

San Jewell

unread,
Aug 4, 2026, 2:02:17 PM (2 days ago) Aug 4
to Biociphers
Hi Xiao, 

I apologize for the delayed response, something has changed about google groups recently and I've not been getting topics sent to me as they used to. 

In any case, if you are looking for alternative 3 prime splicing in particular, I would highly recommend running the "voila modulize" command over your deltapsi or heterogen analysis. It takes the deltapsi or heterogen voila file, the splicegraph file, and .sgc files that cover at least all of the experiments in the voila file. It provides a specific tsv output file per event type, including alternative 3 prime. Please post back here if you have any trouble running it, or questions about the results. Also note that the default behavior for inputs of deltapsi or het is to only output "changing" events (there are many thresholds you can see via voila modulize --help to change how "changing" is detected) ; if you want to see all alternative 3 prime events, you should specify the flag "--show-all" to voila as well. 

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