Table 7.

Distribution-based Algorithm

Data: genes;	/* expression values */
Data: functions	/* for each function */
Result: significance, tailGenes;	/* vector of zeros */
1 normGenes = normalize(genes);
2 hist = zeros(1, nPts);
3 foreach f ∈ function do
4    subset = findPoints(normGenes,f);
5    foreach x ∈ subset do
6       dens = NumberOfNeighbors(x);
7       hist(dens)++;
8    if NunmberOf(genes) greater than a threshold then
9       randHist = findTheoreticalHistogram(1, nPts, normGenes);
10    else
11       randHist = findRandomHistogram(1, nPts, normGenes);
12    significance(f) = chiSquaredGoodnessOfFit(hist, randHist);
13    tailGenes(f) = findTailGenes(hist, randHist);
14 return significance, tailGenes