I have been using rMATS for quite some times, and I found it one of the best tools for splicing studies. However, I don't understand how rMATS handles following situation:
How differential splicing between two samples is calculated for those cases where the alternative splicing has read supports only in one of the two samples of comparison. For example, in many circustances, mutant sample contains additional splice junctions (with several read supports), but that are entirely absent in wild type or control. How rMAT handles those cases. These cases might also have much biological relevance.
To my understanding, rMATS performs the analysis of differential splicing for only those events which are read suppported in both the sample of comparison, rest are not attempted.
Could some body help me understanding this problem.