ratiometricBinomRequiredSampleSize: Calculates the smallest sample size to detect driver genes...

In KarchinLab/cancerSeqStudy: Power and False Positive Analysis of Cancer Sequencing Studies

Description

Effect size is measures as the fraction of sample/patient cancers with a non-silent mutation in a driver gene above the background mutation rate.

Usage

ratiometricBinomRequiredSampleSize(p, desired.power, possible.samp.sizes, mu,
  effect.size, Df = 1, signif.lvl = 5e-06, Leff = 1500 * 3/4)

Arguments

p	the background fraction of total mutations represented by the ratio-metric feature (e.g. inactivating mutations / total)
desired.power	A floating point number indicating desired power
possible.samp.sizes	vector of possible number of cancer samples in study
mu	mutation rate per base
effect.size	fraction of samples above background mutation rate
Df	fraction of driver mutations that are the specific one of interest
signif.lvl	significance level for binomial test
Leff	effective gene length of CDS in bases for an average gene

Value

List containing the smallest effect size with sufficient power

KarchinLab/cancerSeqStudy documentation built on May 8, 2019, 4:47 p.m.