Null and alternative hypothesis being tested in SAM

Question

[r.monil]

Hello all,

I come from a statistical background and I recently started doing work for a bioinformatics research centre. I was wondering if someone could clarify what the null and alternative hypotheses are being tested in SAM (significance analysis of microarrays) in terms of two-sided, one-tailed up, and one-tailed down. Particularly, I'm working on csSAM (cell specific SAM).