**10**wrote:

Hi All,

I am trying to compare 4 groups with 2 samples each (replicates) to understand the differences among each group for a set of 10,000+ genes from a micro array project. I am using the method of permutations and calculating an F statistic as the sample size is too small for a standard t test.

Using an F-statistic and a null of no mean difference, I was able to calculate the obs F- value for each gene in R. But when doing permutations, I understand that there will be 8C2 * 6C2 * 4C2 *2C2 different pairwise permutations which need to be accounted for. I am unable to code for these 2520 permutations in R.

Is there a package that I could look for in R or SAS macro that would help me accomplish the EXACT ORDER of the permutations? I tried the allPerms function from the 'permute' package and the permn function from the 'cominat' package but I am not getting the restricted permutations of interest.

Eg. for 2 group (G1 and G2) setting with 2 replicates each (S1, S2 and S3 , S4): G1 G2 S1 S2 S3 S4 P1: S1 S2 S3 S4 P2: S1 S3 S2 S4 P3: S1 S4 S2 S3 P4: S2 S3 S1 S4 P5: S2 S4 S1 S3 P6: S3 S4 S1 S2

I want to get the exact order of permutations in a 4 group with 2 samples each setting.

Thanks!

**20**• written 4.1 years ago by manali.rupji •

**10**