Statistics / Test proportions

Description

Tests the null hypothesis that two proportions are equal. The test only apply to oportions in dichotomized data.)

Parameters

Details

Tests the null hypothesis that two proportions are equal. The test only applies to dichotomised data, where the data can only assume one of two values. Unless the raw data imported into Chipster already contained dichotomised data the tool Preprocessing / Dichotomise contionious data will have to be applied to the the data prior to using this tool.

The result file adds two columns to the input data table, one with the raw p-values from the test of equal proportions and one with the same values subjected to multiple testing correction. The smaller the value the smaller the probability that the tested proportions are equal.

Multiple testing correction options are Bonferroni, Holm, and Hochberg for family-wise error rate (FWER) and Benjamini-Hochberg, Benjamini-Yakutieri and Storey-Q for false discovery rate (FDR). Of these Bonferroni is the most conservative, returning the smallest number of significant results, and FDR-based adjustments are less conservative, and return more significant results.

Output

A text file containing the gene expression values plus the raw and adjusted p-value for the test.

References

This tool uses Bioconductor packages multtest and qvalue.