Table 1 Properties of the four category tests
From: FUNC: a package for detecting significant associations between gene sets and ontological annotations
Test | Gene associated variable | Alternative hypothesis (one side) | Alternative hypothesis (other side) | Example |
|---|---|---|---|---|
Hypergeometric | binary | Category contains lower proportion of variable "1" as the root category | Category contains higher proportion of variable "1" as the root category | Gene list with detected genes on an array; "0" is not differently expressed, "1" is differently expressed |
Wilcoxon rank | continuous | Sum of ranks of genes in the category is higher than all other genes | Sum of ranks of genes in the category is lower than all other genes | Gene list with detected genes on an array; continuous variable is the probability for being not differently expressed |
Binominal | two counts | Frequency of countA in category is lower than in root category | Frequency of countB in category is lower than in root category | CountA is amount of SAGE tags in experiment A, countB is amount of SAGE tags in experiment B |
2 × 2 contingency | four counts | Counts are dependent and countA/countB < countC/countD | Counts are dependent and countA/countB > countC/countD | CountA and countC are differences at nonsynonymous sites between and within species, countB and countD are differences at synonymous sites between and within species, respectively |