**100**wrote:

I have around 200 subjects and I would like to check the power for 90 of those subjects. It is a case control study.

I did this:

```
fc <- function(x){exp(rnorm(x, log(2), 0.5*log(2)))}
#fc=2
size1 <- ssizeRNA_single(nGenes = 544, pi0 = 0.8, m = 90, mu = 3489, disp = 0.17, fc = fc, fdr = 0.05,power = 0.8, maxN = 120)
size1$ssize
p=size1$power
p=as.data.frame(p)
p[p$n==90,]
```

I got this:

```
p[p$n==90,]
n 0.8
90 0.949
pi0 ssize power
0.8 17 0.8
```

Plot is attached. Does this mean that to achieve power of 80% I wold need 17 cases and 17 controls? Or to achieve power of 90% I would need 90 cases and 90 controls?

Does it make more sense to fix fc=2 in that case I would be to detecting log fold changes of at least 2 or greater given FDR=0.05 as opposed to detecting log2 fold changes across the entire fold-change distribution (as I did above)? Is this a paired analysis for power/sample size?

**50**• written 12 days ago by anamaria •

**100**