Table 4 2 × 2 Table for testing whether pathway j was over-represented in the identified genes (To test H₀: the number of genes in pathway p is independent of the number of identified genes relevant to CAN. Vs. H₁: the number of genes in pathway p is over-represented in the identified genes relevant to CAN, since all the marginal values are given, Fisher's exact test is used. The p-value from the test is $P(G_{ID,p}\ge g_{ID,p})={\displaystyle \sum _{t\ge g_{ID,p}}\frac{\left(\begin{array}{c}{G}_{ID}\\ t\end{array}\right)\left(\begin{array}{c}{G}_0-G_{ID}\\ G_{0,p}-t\end{array}\right)}{\left(\begin{array}{c}{G}_0\\ G_{0,p}\end{array}\right)}}$, where g_{ID, p}denotes the observed value of G_{ID, p})

From: A non-parametric meta-analysis approach for combining independent microarray datasets: application using two microarray datasets pertaining to chronic allograft nephropathy