In the EA sample, 347 categories/pathways were observed to be enriched at p<0.05, 96 at p<0.01 and 15 at p<0.001. These numbers are significantly larger than expected by chance (p=0.003, p=0.004, p=0.009 respectively), suggesting a biological basis for AD captured by the GO and KEGG annotations. When gene set pruning was applied, the excess of enriched categories became even more significant (Table 4), suggesting that the results are not being driven by a few well-annotated areas of biology, each containing several closely-related GO categories/KEGG pathways.
