Hello everyone,
I want to make a power calculation for a miRNA RVAS using the SKAT-O package. The analysis is for a single miRNA where I identified 3 rare variants (MAF<0.01) in my patient cohort (n=6/947) and in no controls (n=0/755). Script requires to set the "subregionlength", which is based on the average length of the gene(s) of interest. The average genomic length for miRNAs is, at best, around 150nt if we take into account flanking regions that could impact cleavage steps and maturation.
If I use this:
power_test<-Power_Logistic(SubRegion.Length=150, Prevalence=0.01, Case.Prop=0.69, Causal.Percent=60, Causal.MAF.Cutoff=0.01, alpha=c(0.05,10^(-3),10^(-6)), N.Sim=1000, OR.Type= "Log", MaxOR=7, Negative.Percent=10)
I get this error:
Error in base::colSums(x, na.rm = na.rm, dims = dims, ...) : 'x' must be an array of at least two dimensions
If I adjust the SubRegion.Length=500, I get :
Get_RequiredSampleSize(power_test, 0.8) $
alpha = 5.00e-02[1] "> 5000"$
alpha = 1.00e-03[1] "> 5000"$
alpha = 1.00e-06[1] "> 5000"
which is, for our capacity, unreachable (total cohort size currently = 1702).
Would you know if there is/are any alternative(s) to the script usage or other aspects?
Ty in advance, Alex
