Question regarding voila modulize output

[chris chris]

Hello,

I just have a few questions regarding the output tsv file from the "voila modulize" command.

I've attached one of the output files that I have for alternative last exons.

The top three entries in this output (gene-LEPR_1 module, ordered by probability changing) have identical values in all columns except for the "event_id" and "junction_name". The two events "gene-LEPR_1_ale_1" and "gene-LEPR_1_ale_3" among these three entries are both classified as "distal" as well, which make these two events identical except for their event_ids.

If all values other than the event_id are identical, is there a reason why these are considered separate events by voila? This is one out of a couple of instances I've noticed in the output.

Also, what coordinates should I be using in this output tsv file to determine the end of the proximal exon and the end of the distal exon? Thanks.

Best,

Chris

[chris chris]

You can ignore the question above.

Is there a page that explains the output format from the modulize command?

There are hyperlinks at "https://biociphers.bitbucket.io/majiq-docs-academic/modulizer/output.html" but they don't seem to lead to anything. I'm asking this question because some results have empty values for the "probability changing" columns.

Thanks.

[San Jewell]

Hi Chris,

I'm not sure what happened but it seems like my reply vanished. If you get two messages, my mistake.

Anyway, is this a dpsi or het analysis? And could you please show one row of the TSV where the problem happens? I believe there are a few cases that can cause the value to be undefined but the most common is that the lsv/junction is not quantified in enough experiments.

-San

[chris chris]

Hi San,

This is from a dpsi analysis. I've attached the event where this happens (multi_exon_spanning.txt), and I guess this also happens because the event is classified as de novo. This is ok, since I'm not working with de novo events at the moment.

I have a different question regarding the cassette.tsv file output from modulize though. (i've attached the file as well.)

Lines 2 and 4 are pointing to the same coordinates according to the "reference_exon_coord", "spliced_with_coord", and "junction_coord" columns, yet these two lines have different "probability_changing". What does the probability_changing value mean here?

Thanks.