Power Analysis - Sample Size Calculation

[jroussarie]

Hi all,
We have been asked by reviewers of a grant application to provide them with a power analysis, to determine how many biological replicates we were including in our RNAseq experiment.
While there are resources available to easily do sample size calculations for microarray experiments by uploading a pilot experiment, I guess RNAseq is too new for the equivalent to exist.
Does anybody have tools to do that, clues or thoughts on the topic?
Thanks very much in advance!

[Drerio]

Hi jroussarie,
We are wondering the same thing. Have you had any luck with the approaches described here?
Thanks!

[Rfriedman]

Power Analysis - sample size calculation

Dear Forum,

I would like to withdraw my previous post on the subject.
The best way I can figure is:

1. log2 transform number of counts,logcounts=log2(counts+1), for
each replicate of "typical" transcript.
2.Based on these numbers perform power/sample size calculation as
for a t-test. use p=.001 to compensate approximately for multiple testing.

With hopes that this helps,
Rich