## Permutation - distribution

Dear All,

I'm in need for some help.
I have two data sets looking like this (just as an example):

Dataset 1:
GCA GCT GCC GCG
9.43 7.59 43.7 39.2
8.20 5.3 41.1 37.9
... (an additional of 2000 rows of data but they are NOT replicates)

Dataset 2:
GCA GCT GCC GCG
18.3 17.3 32.5 31.8
... (an additional of 2000 rows of data but they are NOT replicates)

My problems are:
I would like to test if the average number for GCA in dataset 1 is different from that the average number for GCA in dataset 2. I would like to do this as follows:
1. combine both datasets (no help required)
2. divide the total dataset (so combined) randomly in 2 datasets and determine the average for, e.g. GCA and I want to do this 100,000 times. (--> request some help)
3. These 100,000 averages will used to create a null distribution (--> some help needed)
4. Finally, i would test if the average value of data set 1 falls under the distribution range of null distribution (--> help needed)

hopefully can somebody can aid me with this statistical analysis.

Best wishes,