Randomization Tests

To confirm validity of data sets, randomization tests are often performed usually with the aid of computer programs. These randomization tests permit an investigator to ask the question if the data were collected differently, could the overall results be significantly different. Permutation tests, such as the 'exact test' shuffle the original set of genotypes obtained in a population database to examine how unusual the original sampling of genotypes is. This shuffling generates a new genotypic distribution that can be compared to the original one.

Re-sampling tests referred to as 'bootstrapping' or 'jack-knifing' can be performed on data sets as well. Bootstrapping is a computer simulation where the original n observations are re-sampled with replacement. Jack-knifing on the other hand involves re-sampling by leaving one observation out of the original n observations to create n samples each of size n — 1. Most papers in the literature describing population data sets utilize the exact test with 2000 shuffles although some have reported shuffling as many as 100 000 times.

Since we are only sampling a DNA profile one time in most cases we perform statistical tests to estimate expected variability if the test were performed again. In the end, a number of statistical tests are performed on genetic data to estimate genotype frequencies since many genotypes are rare and may not be seen in population samples tested (see Chapters 20 and 21).


Population genetics is the study of inherited variation and its modulation in time and space. It is an attempt to quantify the variation observed within a population group or among different population groups in terms of allele and genotype frequencies. Great genetic variation exists within species at the individual nucleotide level. For example in humans, approximately 10 million nucleotides can differ between individuals (see Chapter 2). The genetic difference between individuals within human racial groups is much greater than the average difference between races (Barbujani et al. 1997).

As discussed in Chapter 2, diploid individuals have two copies of each autosomal gene (or DNA marker): one of paternal origin (sperm) and one of maternal origin (egg). If the alleles obtained from the sperm and the egg differ, then they are termed heterozygous for that locus whereas if the individual received two identical alleles from both parents they are described as being homozygous.

Variability within a locus has to be stable enough to accurately pass the allele to the next generation (i.e., possess a low mutation rate) yet not be too stable or else only a few alleles would exist over time and the locus would not be as informative (i.e., useful in human identity testing applications). The simplest description of variation is the frequency distribution of genotypes. A measure of this variation is the number of heterozygote individuals present in a population.

Population genetic forces including mutation, migration (gene flow), natural selection, and random genetic drift all affect gene frequency of alleles present in a population. Over time, isolated populations diverge from one another, each losing heterozygosity due to inbreeding. The gene selection pool is smaller in isolated groups and therefore, not as much shuffling of genes exists.

Was this article helpful?

0 0
Prevent Stuttering

Prevent Stuttering

Are You Suffering From Social Withdrawal? Do People Shun Or Ostracize You Because You Have A Hard Time Getting Some Of Your Words Out? Or Does Your Child Get Teased At School Because They Stutter And Cant Speak Like Everyone Else? If you have answered yes to any of the above, then you are in the tiny percentage of people that stutter.

Get My Free Ebook

Post a comment