This hypergeometric analysis indicates that the 4.1 fold greater enrichment in the RAG/CsNp genes was highly significant (p = 0.000015) whereas there was no overall change in growth factor receptor genes (-1.1 fold, p = .42) and significantly fewer genes were found in the apoptosis gene group than expected by chance alone (-2.2 fold, p = .02).
This array was chosen because it contained pre-validated primer sets for three reference genes (Rplp1, Rpl13a, and Ldha) that were unchanged in our array data, six genes that showed a Benjamini-Hochberg corrected significant effect of treatment (Actb, Hspb1, Gfra1, Fgf9, Lifr, and Stat3), as well as several other neurotrophin-related genes that showed a nominally significant effect of treatment (e.g., Gfra2, Ngfrap1, and Npy) that did not survive FDR correction (see Methods section).