We determine whether the differentially expressed genes are associated to a chemical by assessing if the expressed genes are enriched for a chemical-gene set, or contain more genes from the chemical-gene set than expected at random using the hypergeometric test.
We then tested whether a cluster contained more of a particular biological relationship than would be expected by random chance using the hypergeometric distribution with a modified Bonferroni correction as implemented in GeneMerge [25].
A p value is computed as the probability of obtaining an intersection at least as large as the one observed by random sampling using the hypergeometric distribution.
Next, the overlap between genes with high confidence hits of the site in their upstream regions and the set of genes expected to be regulated by the site was assessed by estimating the probability of obtaining the observed amount of overlap (enrichment) by random chance using the hypergeometric distribution.
The most representative functional annotations for each experimental condition are identified using the hypergeometric distribution to determine the probability of random occurrence of functional terms (functional enrichment).
The probability of finding r out of k ORFs with a specific transcription factor (TF -motif pair (or a combinaTF -motifeof) within a class by random sampairg is calculated using the hypergeoretric sampling distribution.
Indeed, the mean number of targets involved in ion transport for 39 random miRNAs (250 genes) gave a P-value of 5.97 × 10− when tested using the hypergeometric distribution.
Using the hypergeometric distribution analysis, we calculated the probability that the proportion of SUMO substrates mapped to a given category could occur by random chance, given the fraction of all fly proteins that map to that category (Table S4, S5).
Fisher's exact test reports the exact p value under a null hypothesis of independent categories using the hypergeometric distribution.
The GO categories and KEGG pathway enrichment analysis of DEGs was performed using the hypergeometric distribution test compared to the whole-transcriptome background.
