I am interested in using either RAD or GBS method to find the variations between 1 population of susceptible insect and 3 population of resistant insect. The insect genome has not sequenced yet, but I assume that the genome size would be about 400-500 MB. How many samples from each population I need to get? Can I get individual insect from each colony and consider it as a sample or all samples need to have different genotypes? Is there any advantage for either of these methods?