Determining RNAseq background noise

Question

[marcos.georgiades.18]

Hi,

I am interested to see if read counts of three RNAseq samples (same tissue) are above background noise. I was wondering whether there is a way to approach this systematically rather than arbitrarily setting a count threshold and remove transcripts below it. For example would it be possible to determine a distribution for the background noise and compare each transcript's counts against it?

Many thanks:), Marcos