Mikado compare output description, clarification needed

[sanyalab]

Hi,

I am using the compare module to compare transcript alignment files to the genome annotation. I wanted to know why you output the Recall and Precision for the best match to the reference transcript at a locus, instead of an average or median or some other statistic that is more representative of the support for a reference transcript. I'll state an example.

Say you have a reference transcript with three exons and 2 introns. At that locus 10 transcripts align. 9 of them support one intron only, and just one supports 2 introns. On running the code, you will report the recall and precision of the one transcript that matches both introns. However, 9/10 transcripts support just one intron. Why not report the median/average precision and recall values of those?

Thanks
Abhijit