les (version 1.22.0)

threshold,Les-method: threshold

Description

The 'threshold' method estimates a suitable threshold $Theta$ from the data. Thresholding provides the ability the define distinct Loci of Enhanced Significance (LES) with the 'regions' method in a later step.

Usage

threshold(object, grenander = FALSE, verbose = FALSE, ...)
"threshold"(object, grenander = FALSE, verbose = FALSE,
...)

Arguments

object

Object of class 'Les' as returned by 'estimate' or 'Les'.

grenander

Logical indicating whether the Grenander estimator for the cumulative density should be used (default: FALSE). For details see below and at the 'GSRI' package.

verbose

Logical indicating whether a summary of the estimated number of regulated probes should be printed on screen (default: FALSE).

...

Further arguments passed to subsequent functions.

Value

Object of class 'Les' with additionally filled slots:

nSigProbes

Estimated number of significant probes.

theta

Estimated threshold $Theta$.

Details

This method estimates the number of probes with a significant effect $R$. The estimation is based on the p-value distribution. The analysis is based on the 'Gene Set Regulation Index' by Bartholome et al., 2009.

Estimation of the threshold is independent of the computation performed by the 'estimate' method.

A reasonable estimate for the cutoff value $Theta$ can be chosen such that $|Lambda >= Theta| = R$.

The Grenander estimator for the cumulative density results in a conservative estimate for the number of significant probes with decreased variance.

A reasonable subsequent step is to call 'regions' to find distinct Loci of Enhanced Significance.

References

Kilian Bartholome, Clemens Kreutz, and Jens Timmer: Estimation of gene induction enables a relevance-based ranking of gene sets, Journal of Computational Biology: A Journal of Computational Molecular Cell Biology 16, no. 7 (July 2009): 959-967. http://www.liebertonline.com/doi/abs/10.1089/cmb.2008.0226

See Also

Package: les-package

Class: Les Methods and functions: Les estimate threshold regions ci export plot

Examples

data(spikeInStat)

x <- Les(pos, pval)
x <- estimate(x, 200)
x <- threshold(x, verbose=TRUE)