Entering edit mode
Michael Salbaum
▴
80
@michael-salbaum-5309
Last seen 10.2 years ago
Dear All,
I wonder if someone could point me towards some guidance using the
package 'sizepower'.
The archives have not netted much in this regard; the vignette was
very useful.
I'm doing a prospective power analysis for an array experiment with a
matched pair design.
I found the function power.matched(ERO, G0, absMu1, sigmaD, n) useful,
however, I have a few questions.
Two issues:
1. I don't have a good feel for the parameter ER0 - mean number of
false positives.
Of course I'd like it to be as low as possible, so I simply ran
scenarios with setting ER0 at numbers we'd be able to tolerate and
calculate power/sample size accordingly.
2. Pardon my ignorance, but I'm also struggling with parameter G0 -
anticipated number of genes in the experiment that are not
differentially expressed.
At the moment, I'm assuming G0=G i.e. number of gene (probes) tested.
Does G0 need to be based on the number of gene probes on the array
(~46,000)? Inspecting a similar paradigm on the same array platform on
GEO reveals that only about ~12,000 gene probes are consistently
detected. I'm aware that it doesn't make a huge difference (and almost
none in practical terms), but I'm curious.
I've left the parameter absMu1 at 1 (as I understand it, this
represents effect size, as fold change of at least 2).
My sigmaD parameter comes from a preliminary experiment: gene
expression by qPCR, standard deviation of the (log2) differences
between pairs.
Thanks for the help.
Cheers, michael
P.S. If I should be using a different package for the task at hand,
I'd like to hear such a recommendation as well - thanks.
R version 2.15.1 (2012-06-22) -- "Roasted Marshmallows"
Copyright (C) 2012 The R Foundation for Statistical Computing
ISBN 3-900051-07-0
Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
Natural language support but running in an English locale
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
[R.app GUI 1.52 (6188) x86_64-apple-darwin9.8.0]
[Workspace restored from /Users/jms/.RData]
[History restored from /Users/jms/.Rapp.history]
> library(sizepower)
> power.matched(1, 12000, 1, 0.731, 20)
$power
[1] 0.9854905
$psi1
[1] 37.42788
> power.matched(1, 46000, 1, 0.731, 20)
$power
[1] 0.9693688
$psi1
[1] 37.42788
>
J. Michael Salbaum, Ph.D.
Associate Professor
Pennington Biomedical Research Center
Louisiana State University System
6400 Perkins Road
Baton Rouge, LA 70808
(225) 763-2782
[[alternative HTML version deleted]]