voila tsv outputs

[Poorva Kumari 20204]

Hi!

I ran voila tsv and saved all the output tsv files generated from the splicegraph.sql and deltapsi file. The output files contains tsv files for the individual events like alt5prime.tsv, ale.tsv, cassette.tsv etc, but can you pls explain what's the difference between these individual events' files and the summary.tsv/junctions.tsv/heatmap.tsv files/other.tsv? Also the voila view command corresponds to which of these tsv files? 

Lastly, what's the difference between alt5prime.tsv and p_alt5prime.tsv, alt3prime.tsv and p_alt3prime.tsv?

Thanks in advance.

[San Jewell]

Hi,

I believe the majority of these questions can be answered by looking at the documentation for the modulizer: https://biociphers.bitbucket.io/majiq-docs/modulizer/index.html

However, to answer specifically:

-summary.tsv has a list of the type and number of events per detected module, but does not contain specific information about each individual event

-junctions.tsv is basically all of the other event type files pasted together, in that it will contain all of the quantification columns for every junction if you don't care about looking for specific event types

-heatmap.tsv is a (junctions)x(samples) matrix, whereby each module is represented by 1 and only 1 junction. The purpose of this file is to allow easy unbiased clustering-style analyses by ensuring you don't have redundant quantifications. In order to select the one junction per module there are two options to choose:  

1. Shortest junction (default if input files are only PSI)
2. Greatest |dPSI| (default if input files include DPSI or HET)

(note these defaults can be overridden by using the "--heatmap-selection" command line option)

-other.tsv is a list of events which did not match the pattern of _any_ of the other event types that we look for. We don't yet know what to call them!

-I am unsure what you mean by "which voila view command corresponds to these files" , all of these outputs are generated by "$ voila modulize" , with "$voila view" being a completely separate program which shows a webview of the results

-The putative alt3 and alt5 prime sites look like they might have been alt3 or alt5 prime, but instead of an exon they fully or partially contain a retained intron in its place. See full detail here: https://biociphers.bitbucket.io/majiq-docs-academic/modulizer/event-types.html#putative-alternative-splice-sites-p-alt-5-3

Let me know if it helps!

Thanks,

-San

[Poorva Kumari 20204]

Hi,

Thanks a lot for answering my queries. By which voila view command corresponds to these files I meant - do the list of genes and their modules shown on the webview results shown after running the voila view command, correspond to any of the tsv files (e.g. the summary.tsv(?) ) which we get after running the voila modulize command?
Also what is the output generated by voila tsv command? How is it different from any of the voila modulize outputs?

[San Jewell]

Heya,

No worries!

Voila TSV is a more legacy command than voila modulize, meant for general usage. It contains many of the same filters but it does not make modules. It is a flat file with all quantifications that pass filters. Lines are shown per-LSV instead of per-junction.

The files of modulize show all genes at once ; whereas voila view tends to drill down to specific genes/lsvs. If you see a module or event of interest in the modulize output, you can open the voila view and enter the LSV or gene into the search box and click on it. It will show the splicegraph for the whole gene and more detailed information about the LSV which can help better understand and "see" the structure that the modulizer found.

Let me know if that makes sense.

-San

[Poorva Kumari 20204]

Hi,

Thanks again for lucidly answering my questions. When I run voila view, among manyI got few genes and their splicing events on the webview results, but I cannot find about them in any of the tsv files which I had generated using voila modulize with the same input files. 

[San Jewell]

Hiya, no worries.

I think this is likely because the modulize filters are too strict. It usually only looks for events of interest. When supplying dpsi or het inputs, by default it will also only look for changing events.

You probably want --show-all , and perhaps --decomplexify-reads-threshold and --decomplexify-psi-threshold set to lower than default values as well. I would recommend looking at the voila modulize --help ; text, to get an idea of all possible filter options.

Thanks,

-San