fishersPvalueSumTest: Combine multiple independent P-values together using Fisher's...

fishersPvalueSumTestR Documentation

Combine multiple independent P-values together using Fisher's method

Description

Fisher's method for combining P-values together consists of studying a test statistic of T = ∑_i -log(P_i) and finding the corresponding point in the cumulative distribution of an appropriately parameterised chi-squared distribution.

Usage

fishersPvalueSumTest(testPvalues, ...)

## S4 method for signature 'numeric'
fishersPvalueSumTest(testPvalues, ...)

## S4 method for signature 'Pvalues'
fishersPvalueSumTest(testPvalues, na.rm = TRUE, rate = 1, ...)

Arguments

testPvalues

A numerical vector of P-values to be tested.

...

extra-arguments

na.rm

remove NA. Default set to TRUE

rate

Provide the rate for the exponential in the sum (distributed as a gamma)

Details

The basic assumption is that P-values are uniformly distributed under the null-hypothesis, meaning that -log(P) \sim exp(1). Recall that if X \sim exp(1) then ∑_{k} X_{k} \sim Γ (k,1) \sim χ^2 (2k) (note that the gamma and erlang distributions are identical for integer k). We therefore simply calculate the point in the cumulative of the gamma distribution

Note that the use of this method for analysis of omics datasets was pioneered by Luo et al.

Value

combinedPval A single P-value combining the results of multiple independent hypothesis tests into one

Methods (by class)

  • numeric: Convert to P-values on the fly

  • Pvalues: Fisher's combined probability test

References

https://en.wikipedia.org/wiki/Fisher%27s_method

https://en.wikipedia.org/wiki/Erlang_distribution

Luo, W., Friedman, M. S., Shedden, K., Hankenson, K. D., & Woolf, P. J. (2009). GAGE: generally applicable gene set enrichment for pathway analysis. BMC Bioinformatics

See Also

betaUniformPvalueSumTest

Other P-value sum tests: betaUniformPvalueSumTest()


adamsardar/metaDEGth documentation built on Sept. 28, 2022, 10:12 p.m.