fdr.optimization: Function for numeric optimization to get an FDR corrected...

In heiniglab/scPower: Experimental design framework for scRNAseq population studies (eQTL and DE)

Function for numeric optimization to get an FDR corrected significance thresold

Description

Function for numeric optimization to get an FDR corrected significance thresold

Usage

fdr.optimization(
  x,
  fdr,
  m0,
  type,
  exp.vector,
  es.vector,
  nSamples,
  mean.vector,
  disp.vector = NULL,
  useSimulatedPower = TRUE,
  simThreshold = 4,
  ssize.ratio.de = ssize.ratio.de
)

Arguments

x	Adjusted significance threshold to be found during optimization
fdr	Chosen FDR threshold
m0	Number of null hypothesis
type	Either DE or eQTL power
exp.vector	Vector of expression probabilities for each DEG/eQTL
es.vector	Effect size vector (Fold Change for DEGs, Rsq for eQTLs)
nSamples	Sample size
mean.vector	Mean value for each DEG / eQTL gene
disp.vector	Dispersion value for each DEG (only required for DE power)
useSimulatedPower	Option to simulate eQTL power for small mean values to increase accuracy (only required for the eQTL power)
simThreshold	Threshold until which the simulated power is taken instead of the analytic (only required with eQTL power)
ssize.ratio.de	In the DE case, ratio between sample size of group 0 (control group) and group 1 (1=balanced design)

Value

Optimizing this function to 0 will lead to the correct adjusted significance threshold x

heiniglab/scPower documentation built on Jan. 9, 2025, 12:13 p.m.