predictPower: Estimate Power Under Increasing Sample Sizes

Description Usage Arguments Value See Also Examples

View source: R/predictPower.R

Description

Simlilar to estimatePower, power estimations are performed under multiple increasing sample sizes

Usage

1
2
3
4
5
6
predictPower(inputObject, groupVec, isLogTransformed = FALSE,
  dataType = c("RNASeq", "Proteomics"), enableROTS = FALSE,
  paraROTS = list(B = 1000, K = NULL, paired = FALSE, a1 = NULL, a2 = NULL,
  progress = FALSE), minLFC = 0.5, rangeSimNumRep = NA, alpha = 0.05,
  ST = 100, seed = 123, parallel = FALSE, BPPARAM = bpparam(),
  showProcess = FALSE, saveResultData = FALSE)

Arguments

inputObject

a numeric raw Proteomics abundance data matrix, in which rows correspond to proteins and columns correspond to samples.

groupVec

a vector indicating the grouping of samples

isLogTransformed

logical; logical; set to TRUE, if the input data is log transformed.

dataType

"RNASeq" or "Proteomics" indictes the data type of the input data matrix.

enableROTS

logical; if TRUE, Reproducibility-Optimized Test Statistic (ROTS) will be used as the statistical model.

paraROTS

a list object containing addtional parameters passed to ROTS (if enabled), see ROTS.

minLFC

LFC threshold

rangeSimNumRep

a vector of sample sizes under which power will be estimated

alpha

controlled false positive rate.

ST

the number of simulations of abundance data generation and repeated times of statistical test for each protein (>=100 recommended).

seed

an integer seed for the random number generator.

parallel

logical; if FALSE parallelization is disabled; if TRUE, parallelize calculations using BiocParallel.

BPPARAM

an optional argument object passed bplapply to indicate the registered cores, if parallel=TRUE.

showProcess

logical; if TRUE, show the detailed information of each simulation, used for debugging only.

saveResultData

logical; if TRUE, save the simulated data into RData with name pattern "simulated_Data_numRep_X_numSim_XXX_XXXXX.RData".

Value

a list of power predictions for each sample size, grouped in comparisons between each two groups

See Also

estimatePower estimate power based on actual data

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
# Example 1: a random generated Proteomics dataset (10 DE, 100 non-DE)
data(exampleProteomicsData)
dataMatrix <- exampleProteomicsData$dataMatrix
groupVec <- exampleProteomicsData$groupVec

# Run estimation
# Note: Simulation times(ST) is specified as 5 for shorter example runtime
#       For better performence, ST > 50 is recommended
predictedPower <- predictPower(dataMatrix, groupVec,
                               isLogTransformed=FALSE,
                               dataType="Proteomics",
                               minLFC=0,
                               rangeSimNumRep=c(5, 10, 15),
                               alpha=0.05, ST=5, seed=123)

PowerExplorer documentation built on April 28, 2020, 9:03 p.m.