GOTests: Gene set tests statistics
In hxin/topOnto: ontology enrichment analysis

Description Usage Arguments Details Value Author(s) See Also Examples

Methods which implement and run a group test statistic for a class inheriting from groupStats class. See Details section for a description of each method.

GOFisherTest(object)
GOKSTest(object)
GOtTest(object)
GOglobalTest(object)
GOSumTest(object)
GOKSTiesTest(object)

object

An object of class groupStats or decedent class.

GOFisherTest: implements Fischer's exact test (based on contingency table) for groupStats objects dealing with "counts".

GOKSTest: implements the Kolmogorov-Smirnov test for groupStats objects dealing with gene "scores". This test uses the ks.test function and does not implement the running-sum-statistic test based on permutations.

GOtTest: implements the t-test for groupStats objects dealing with gene "scores". It should be used when the gene scores are t-statistics or any other score following a normal distribution.

GOglobalTest: implement Goeman's globaltest.

All these methods return the p-value computed by the respective test statistic.

Adrian Alexa

groupStats-class, getSigGroups-methods

data(ONTdata)
gene.universe <- genes(GOdata)
go.genes <- genesInTerm(GOdata, 'DOID:0050709')[[1]]
sig.genes <- sigGenes(GOdata)
my.group <- new("classicCount", testStatistic = GOFisherTest, name = "fisher",allMembers = gene.universe, groupMembers = go.genes,sigMembers = sig.genes)
contTable(my.group)
runTest(my.group)