I have two populations, CEU and CHS, and 50 individuals for each populations.

I split each chromosome into segments with 100kb window size.

For each individual, there are some some bins(100bp) in each window.

Now, I hope to know, for each window, whether the total number of bins is significant difference between two populations or not.

In such case, which hypothesis test is best ?

I think, in this case, the mean value is not very meaningful, so I do not choose two samples t test.

Is there any other better methods?