Off topic:How to inspect novel antisense based on the distribution of the known natural antisense and those from the expirement

Question

[M K]

Hi All,

I did an experiment to detect novel antisense, and I got the sense and antisense count read counts, then I calculated the expression level for this data. After that, I removed the known antisense from my data and keep it as training sample, and I plotted histograms for both dataset. Now I want to identify the novel antisense by comparing the two histograms so what is the best way to do that (i.e how to set a threshold to do that based on the two graphs)