Moccasin correction/ Voila visualization

[Alessandra Longo]

Dear MAJIQ team,

First of all, congratulations on the release of version 3! It’s great to see that it now includes correction with Moccasin — a very exciting and much-appreciated update.

I had a quick question: after applying the correction, does the visualization in Voila now reflect the adjusted read counts?

Also, I would like to ask in general if it is possible to perform a comparative analysis between two treated groups and their respective controls using this new version.

Thanks in advance for your time and for the continued improvements to MAJIQ

Alessandra

[San Jewell]

Hi Alessandra,

With respect to your first question, voila does not show adjusted read counts in v2 or v3. The read counts shown above splice junctions in voila are median raw read counts. The way moccasin works is by adjusting the bootstrap read rates which are then used to calculate psi. So at the moment 1) bootstrap read rates are not shown in voila and 2) raw read counts aren't adjusted by moccasin. In theory, one or both of these could be changed, but after having discussed it with the lab there was no immediate need for these features. If you would find either one useful, please post here and I will bring it up for consideration.

For the other question "is it possible to perform a comparative analysis between two treated groups and their respective controls" -- This statement is a little bit vague to be answered confidently. In the general use of majiq CLIN, a set of controls is tested against different sets of cases. What approximately did you have in mind?

Thanks,

-San