huge spike in intron retention events between Majiq 1.1 and 2.0

[Bina Sugumar]

Hi, 

Thank you so much for this great tool, its been a pleasure to explore all its features!

I have been seeing a huge number of IR events in the 2.0 run of majiq that were not there when I used 1.1. In the 1.1 run I had 56,655 LSVs and 447 of those had IR events.

In this new run I have 38,326 LSVs and 20,864 have IR events. 

Are there additional parameters I should use aside from the default settings in majiq to make the IR event detection more stringent, since it's likely that the majority of these events are not real?

Thank you,

Bina