FindOptPower: search for the optimal pooled design

Description

Perform a grid search over potential design space, and find the predicted power and validity of the designs.

Usage

FindOptPower(cost, sample.size, MAF, OR, error, costPerExp = 18915, costPerPool = 970, costPerX = 300, lower.P = 20, upper.P = 400, lower.N.p = 2, upper.N.p = 200, lower.Xmean = 4, upper.Xmean = 1280, sig.level = 0.05, Number.Grids = 100)

Arguments

cost

cost constraint of the study

sample.size

sample size constraint of the study

MAF

assumed MAF of the variant of interest

assume effect size (odds ratio) of the variant of interest

error

assume sequencing error rate

costPerExp

cost per experiment

costPerPool

cost per pool

costPerX

cost per 1X coverage

lower.P

lower bound of number of pools in potential consideration

upper.P

upper bound of number of pools in potential consideration

lower.N.p

lower bound of number of pool size in potential consideration

upper.N.p

upper bound of number of pool size in potential consideration

lower.Xmean

lower bound of number of coverage per pool in potential consideration

upper.Xmean

upper bound of number of coverage per pool in potential consideration

sig.level

significance level of the statistic test, usually 0.05 for a single test

Number.Grids

number of grids in the search space, preset as 100

Value

cost
sample.size
constraint.set
scenario.set
designs: the potential designs, validity and power

Details

Given the research question and the study constraints, this function calculates the power and validity of all the potential pooled designs.

Examples

Run this code


# # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
######## Example 1: A simple example, with very rough grid points (only 20X20 grid points)
# # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #

##### Find the optimal design
example.1 <- FindOptPower(cost=452915, sample.size=5000, MAF=0.03, OR=2, error=0.01, upper.P=200, Number.Grids=20)

##### assign a directory to store the contour plots
##### with your own choice
proj.Dir <- paste(getwd(), "/hiPOD_examples", sep="")
if(!file.exists(proj.Dir)) dir.create(proj.Dir)

##### Inferences on the optimal designs
PlotOptPower(example.1, save.contour=FALSE, plot.3d=FALSE)
# # snapshot3d(filename = paste(proj.Dir, "/example1_3d.bmp", sep=""))
ShowOptDesign(example.1, 5)