Working with DNA

The following  is based on an extended discussion with Ian Brennan. Ian is a phD student working on the  phylogeny (classification) of reptiles. His website can  be accessed at

There are four types of DNA in animals.

  1. Untranslated (not used for protein synthesis) regions of DNA (also may be called Junk DNA –its DNA that we don’t know much about) very high mutation rate that does not experience the constraining effects of natural selection

  2. DNA within genes that are found in the nucleus
    2.1 Intron DNA – very high mutation rate that does not experience the constraining 
    effects of natural selection

    2.2 Exon DNA – low high mutation rate that does not experience the constraining effects of natural selection (these are currently considered to be the gold by animal

  3. DNA within genes that are found in the mitochondria very high mutation rate relative to nuclear DNA.

When you are looking at fine scale differences (those between members of the same population or closely linked populations) you want to use DNA with a higher mutation rate. When looking at more distantly related populations it is better to look at more slowly evolving sections of DNA, therefore exon DNA.

What process does Ian use to compare different organism’s DNA?

  • The process that is used exposes the DNA (intron or exon depending on if they are comparing at a fine grain or at a larger level) to restriction enzymes. This chops the DNA into fragments of various lengths.

  • Filter (using a process similar to gel electrophoresis) to fragments approx. 300 base pairs in length (Longer sections are hard to sequence). Get rid of big pieces (>400bp). Make the assumption that similar sized pieces are from same region of the genome in the organisms (or species that are being compared).

  • If doing a fine grain study, they will use intron DNA. They will look for sections of the 300 ish base pair lengths. They specifically look for single nucleotides that vary (pint substitution mutations) - these are called SNPs (single nucleotide polymorphisms). They do this 50 000 to 100 000 SNPs across the genome. This is done multiple times (50 000 to 100 000 times) because it is very unreliable to compare one a few regions of the genome. It is much more reliable if the scientist looks at a large number of SNPs across the genome.

  • When looking at more distantly related species they will study SNPS in exon DNA. On these occasions they target 300 to 400 SNPS on these occasions.

For improved accuracy when understanding the makeup of the genome or sections of this they use ‘tiles’. The scientists use overlapping sections of DNA and their computer skills, to understand the code of the chosen section of DNA.

To read more about SNPs in humans go to 

  • The following image shows some work that Ian has done on Eulampris water skink phylogeny.