Poodle Diversity Project - Blog

  • Primer on Genetics Data #2

    For those who want to understand more about the DNA analysis methodology, here are some more details. Some of your eyes may glaze over, which is fine because the gist is in the previous post, but for those who really want to understand the nitty gritty, here it is.

    The two best known ways to determine genes in an individual differ in several ways.

    STRs are “short tandem repeats” and that’s the older technology, the kind that has proved paternity and guilt in courts of law for decades. So it’s tried and true.

    DNA tends to do predictable things, and one of the things it does sometimes is “stutter” in specific areas when it’s replicating to make new cells, like eggs and sperm. Remember that DNA is coded using 4 elements, represented by G, A, C, and T. It's kind of like Morse cod, which uses dots and dashes in patterns, but using 4 materials.

    So if a part of the sequence of DNA is GCATCTCGCA, when making an egg or sperm, the replication process might stutter and make GCATCATCATCTCGCA in the egg or sperm, which is then passed down to the next generation. As you can see, the CAT in the sequence has been repeated three times in the second example.

    This stutter doesn’t change gene function, but it does act as a sort of marker that researches can find - like putting paint on trees in a large forest to mark a trail. Using places on the DNA that are known to have these stutters, or repeats, individuals can be compared. Each slightly different sequence is called an allele, and each place on the DNA is called a locus. To analyse each dog, they find the same place on the DNA and notate the sequence. Each sequence is given a label; UC Davis uses numbers. Doing this at a large number of loci makes it possible to identify each individual because individuals have unique combinations of these alleles. Relatives have some alleles in common. 

    Over generations, these stutters, or STRs, happen. Different families have different STRs at the same places on the DNA, and the longer a population is separated from others, the more likely they are to have different STRs. Sometimes a breed will have only 4 different STRs at a certain locus, and some breeds will have 10. The more STRs, or alleles, the more diversity there is in a population. 

    Randomly bred dogs, like village dogs in primitive communities, tend to have a great deal of diversity. Highly inbred breeds with few founders and a small breeding population have very little diversity. 

    Researchers design “panels,” or the mechanism by which they check the places on the DNA, that use loci that are well spread throughout the genome, so the STRs they use are on as many different chromosomes as possible. That way they get a very good sampling of the DNA in each individual.  This method can positively identify any human on earth because the chances of each human having the exact same markers at precisely the same places (unless he or she has an identical twin) is extremely low. The advantages of these STRs are that they are economical to process and analyze, they use cheek swabs rather than blood samples, and they identify more recent changes to DNA better than ancient ones. This makes them ideal for assessing diversity of many different wild animals, and now dogs and telling populations apart. They have been used for this purpose for many years. 

    The other method of recording is SNPs. SNPs are “single nucleotide polymorphisms” or each individual base pair in a sequence of DNA. Rather than specific repeats, SNP technology records changes in the single letters from one of G, C, A, or T to another one of those four, or additions of single base pairs or subtractions. This is newer technology and like the STRs, it catalogs only a tiny section of the DNA on each chromosome. It can track minute differences in the DNA that STRs cannot track, and so is better for certain tasks like looking for tiny mutations, or telling identical twins apart. It is no better at recording genetic diversity, however. It is also more expensive to do and often requires higher quality DNA samples, like from blood, and is not as practical for the purposes of testing hundreds of dogs.

    As referenced on the VGL website, the 33 loci STR panel was compared specifically to a 170,000 SNP panel and the results were comparable. That is one way they established the validity of the method for this purpose.

    More to come!