The Science Behind Dog Poop DNA Matching

The process of actually matching the DNA is complex and time intensive. Obviously, we can't hold up the two cotton swabs to a light bulb to see if they match.

DNA matching is not as simple or as quick as you see on TV, in fact it is actually quite complex and is the result of years of genetic research and development by commercial companies and universities across the globe. The reliability and accuracy of DNA is well documented and has been proven in court.

DNA Introduction A single strand of dog DNA is made up of billions of nucleotides and even with the variance in breeds, sizes and appearances 99.9% of the DNA will be the same, however, there are significant size variations in particular sequences that have been isolated as known genetic markers. We use 23 of those genetic markers to create an identity for each dog, which is significantly more than the 14 genetic markers used by the FBI in human cases. The reason we use more markers than the FBI is because the DNA in poop is typically degraded and provides less usable DNA than mouth swabs or blood samples. That means we typically have far less than 23 markers to work with in the poop samples. DNA Collection First the DNA sample is collected using a simple cotton swab to gather some buccal cells or the loose skin cells from inside the dogs cheek. Kits designed and manufactured by Mr Dog Poop Inc make the process simple, fast and effective. Our kits account for temperature and humidity variations and can be stored for long periods, several years in fact, even after the initial DNA collection.

DNA Extraction, Amplification and Analysis Next, we need to isolate the DNA strands from the cells collected. The swab is then put through an extraction or lysis and purification process which separates out just the DNA and removes any impurities that would inhibit the analysis. The process involves several steps of heating, cooling, filtering and washing using a centrifuge. Once the DNA is isolated it is quantified to determine if an adequate amount of DNA was collected. The sample is diluted in a cuvette and placed in a spectrophotometer which uses light at a specified wavelength to measure the density of the concentration in the sample. Then to amplify the DNA, forward and reverse oligos nucleotide primers are added which will allow us to duplicate only the genetic markers that are needed for identification rather than the entire strand of those billions of nucleotides. Amplification is done using a thermocycler in a process known as polymerase chain reaction or PCR. The thermocycler slowly heats and cools the DNA for about 4 hours which causes it to replicate over and over until there are billions of duplicate strands of each genetic marker. Of course, even with billions of strands they are so small they are contained in less than a single drop of water. The amplified DNA, in that single drop, is then analyzed using capilliary electrophoresis which uses an electrical charge to draw the DNA samples through hair sized capillaries filled with a liquid polymer. The size of each marker is then determined by the time that it takes to travel through the capillary and is measured by lasers to identify each fragment's size. That complex series of data points are exported to graphing software where they can be reviewed for accuracy.

Below is a sample of the data provided by the capilliary electrophoresis. The upper graph or electropherogram shows the markers collected from one of the saliva swabs and the lower graph is from a dog poop sample.

For each genetic maker there are 2 alleles represented by peaks on the graph, one is inherited from each parent. This genetic map is our digital fingerprint that we compare with other digital fingerprints from other dogs.

The sample image shows only 5 markers however all comparisons are done by constructing full data sets of all 23 genetic markers logged in our database.

In this case it is easy to see that the markers appear at the same points on the x axis and therefore have the same sizes confirming a positive match.

DNA Database Online Of course all of our matches are done automatically by our CRIME LAB® Database and proprietary matching algorithms. We have developed our own software not unlike that used by the FBI which interprets those data points and finds likely matches. However, the actual data is always reviewed by qualified technicians to further validate and confirm each match. Validity & Reliability of DNA Results The process works very well and has been used for decades in law enforcement in criminal cases. The data is extremely accurate and reliable if the tests are run correctly. It is virtually impossible to match a dog with the wrong poop. Of course data can often be inconclusive and we may not always be able to positively make a match with complete certainty. In those cases we consider that a "No Match".

Accuracy vs Probability - Random Match Probability

Most people believe that a DNA match is 100% conclusive and it could have only come from the accused dog. However, that is not true.

In DNA we do not talk about accuracy but rather the probability that the DNA came from a specific subject. So we use "Random Match Probability" which is the overall probability that the DNA might have come from any dog other than the accused. This number is based on all the dogs ever tested and the frequency of each allele size.

A mathematical calculation is provided with each tested sample to show that it is mathematically improbable while not necessarily impossible that any other dog could match all the markers found in the poop sample.

To understand how this works, let us look at a more simplified scenario with just one matching marker.

For example, if we only use the sex marker and determine that the dog responsible was a male.

The probability that the sample came from one particular dog would be around 1 in 2, that is if we assume that half of the population is male and half is female.

Essentially, this could be any one of half the dogs on the property.

As we add more markers, those numbers change and match probability is often one in billions or one in trillions and even higher.

That number is important, because if the Random Match Probability is low, even though we have a match, it is impossible to say with any level of certainty that it is without doubt one particular dog.

For this reason, we require a minimum number of matched markers despite often believing it is one of the registered dogs. We have to consider that any one of thousands of dogs could have entered the property and pooped. So it would be unfair to match a dog just because they are the only dog registered that matched the sample when we cannot rule out other dogs that might have visited the property.

In rare cases, we have seen dogs from the same family matching several markers and if two dogs match the sample with no way to exclude either dog, it would be impossible to blame one dog over the other.

Of course, if they live in the same apartment, with the same owner, we at least know who to fine. But if they have been split up and have 2 different owners on the same property, we would not be able to make the determination without additional markers. In those cases, it may not be possible to get a match with a degraded sample with only a few markers and a much better sample would be required.

In uncertain cases, we typically notify the property and let them know we suspect a certain dog or dogs and advise them to keep a close watch over those dogs.

Again, the DNA is a tool and it is important that we look at all the evidence available to draw conclusions. If the sample has a low Random Match Probability, does not match any other dogs on the property but was found 10 feet from the door of the suspect, there is a pretty good chance the suspect is responsible.

Any additional information can also help narrow down poopetrators, photo's of the sample, location and a knowledge of which dogs are being walked in that area. For example, a Chihuahua will have significantly smaller poop than a Siberian Husky. So while properties often expect the DNA to do the work it never hurts to have additional data.

If this had been a real criminal investigation, we would be getting much more data to work with such as the weight of the sample, the diameter, consistency, temperature, weather conditions, undissolved food content, hair strands, photos etc. Which is another reason why properties cannot expect the same results that would be provided in a much more thorough forensic process.

Our Genotyping Is NOT Breed Testing

Many people believe we can determine the breed of dogs from the DNA sample provided. However, the science of breed testing is still more predictive than actually being scientifically accurate. In fact, the science of canine DNA overall is still very young compared to our level of understanding of human DNA.

While it may be possible to predict the breed of a dog through DNA analysis, this is not something we do in our lab. That would require different resources and tests.

Our focus is purely scientific and only allows us to compare two samples and determine if they came from the same source.

Therefore, we cannot tell anyone what type of dog they have or what mix of breeds the dog may be either from the buccal swab or poop samples.