[BioC] Testing sensitivity and specificity on random biological sequence

[Jiří Hon]

Dear Bioconductor community,
we developed a package for detecting potential intramolecular triplex 
patterns in DNA sequence 
(http://www.bioconductor.org/packages/release/bioc/html/triplex.html) 
and we're going to review the sensitivity and specificity measures of 
the detection algorithm. But we are doubtful what is actually the right 
way to do that, I mean how should we generate random testing DNA 
sequences to get relevant results?

We found some articles citing the uShuffle tool 
(http://digital.cs.usu.edu/~mjiang/ushuffle/) so that might be the 
direction. But we are curious if someone from Bioconductor community has 
some other experiences with solving this problem. We would be really 
thankful for that.

So what tools do you use for this purpose? And what sequence properties 
(k-let counts, etc.) you think should be preserved?

Thank you a lot,
Jiri Hon