TABLE 1.

Basic steps of the algorithm that selects peaks based on their ability to maximize the deviation from the control distribution of expression values

k = 0,S(k) = Ø,D(k) = − ∞, max = − ∞ k = k + 1 h* = arg max N(h), N(h) = number of genes with corresponding hash value h G(k) = {gi; hash(gi) = h*}, the subset of genes that hash to h Evaluate F(Ygi(t)); t = 0, K, T; gi ϵ Σ Evaluate $$D(k)=\underset{t}{max}\underset{g\text{i}\in \Sigma }{max}|F[Y_{g\text{i}}(t)]-F[Y_{g\text{i}}(0)]|$$ If D(k) > max Max = D(k); F = k; Go to (ii) until all peaks have been added For a = 1 to F Select Σ = S(a – 1) U G(a)