杏吧原创

Man鈥檚 best friend is turning into a witness for the prosecution, with canine evidence cracking crimes from dogfighting to murder

IN APRIL 2009, 16-year-old Seyi Ogunyemi and a friend were walking in Larkhall park in south London when they were surrounded by a 20-strong group of youths. Ogunyemi turned to flee over a fence, but he was mauled to the ground by Tyson, a pit bull/mastiff cross. Having set his dog on the teenager, 22-year-old Chrisdian Johnson then took out a knife and stabbed Ogunyemi to death.

Police investigating the gang-related crime found a trail of blood running from the park right to Johnson鈥檚 front door and suspected it belonged to the dog, which had been injured during the attack. When DNA analysis matched this blood and the blood on the victim鈥檚 body to Tyson, the case was clinched. In March, Johnson was found guilty of murder 鈥� condemned by evidence from his own dog, which he had used as a weapon.

This conviction was the first to use reference samples from a new UK database of dog DNA, which opened for business earlier this year. With more cases in the pipeline, such services are clearly in demand. The US is leading the game and has several similar databases that have sprung up over the decade, and some high-profile criminal cases hingeing on animal DNA. This year, a DNA database dedicated to combating dogfighting was launched in the US (see 鈥淏usting the fight club鈥�). Doggy forensics looks set to take off.

The first case in which animal DNA was used to catch a killer involved a pet cat called Snowball. In 1996, a jury at the supreme court of Prince Edward Island, Canada, found Douglas Beamish guilty of murdering his common-law wife, Shirley Dugray. Beamish was implicated when a few of Snowball鈥檚 hairs were found in the lining of a man鈥檚 leather jacket which was stained with Dugray鈥檚 blood ().

Extending this approach to dogs was the obvious next step. 鈥淲hen I used to do a lot of forensic casework we got many cases involving canine material: hair, saliva from bite wounds, and blood,鈥� says geneticist Sree Kanthaswamy, who runs a dog DNA database at the University of California, Davis. 鈥淚n the US you have a dog for every four people,鈥� he adds. With more than 70 million pooches living alongside humans there is a good chance that dog traces will be left at many crime scenes.

As with human DNA matching, profiles from dogs are based on just a few marker regions of the genome. And as with human DNA evidence, the weight that a canine DNA match is given in court will depend on how common the particular pattern of DNA is found to be among dogs generally. To work out the probability that another dog might have the same profile, forensics experts need a large database with as many reference samples as possible. Some believe it is also important that the dogs come from the population you are investigating. 鈥淚f you use a database from a different country it may not be representative,鈥� says Rob Ogden who, along with Ross McEwing of , London, runs the UK鈥檚 new facility, which is based in Edinburgh.

Convincing evidence

The UK database holds profiles consisting of 13 microsatellite markers 鈥� short repetitive stretches of DNA present in non-coding regions of the genome 鈥� and one sex marker from about 400 dogs from Scotland and England, representing 12 breeds and some cross-breeds. Using this database in the Ogunyemi murder case, the team concluded that there was a 1-in-a-billion chance that a dog other than Tyson would have had the profile derived from the blood and saliva at the scene of the crime. In this case, the canine forensic evidence was strong. There was lots of blood at the scene, and Tyson could easily be swabbed for DNA 鈥� in the same way cheek swabs are taken from human suspects. Some other cases are not so straightforward.

When investigators have only a single dog hair or the tiniest trace of canine blood to go on, there may not be enough intact genetic material to run a standard profile. Then, instead of looking for genetic markers from regular DNA, forensic scientists may have to rely on mitochondrial DNA found within the cell鈥檚 power houses, which is present in far greater quantities than nuclear DNA. This approach is less reliable, since there is less genetic variation between dogs in mtDNA than in nuclear DNA. Here, forensic scientists look at the variation in the mtDNA sequence itself rather than looking at markers, but even if there is a complete sequence match there is still a reasonable chance that the mtDNA could belong to another dog.

Some sequences are so common they have a match probability of about 1 in 7. However, around a quarter of the dogs in the world have fairly uncommon mtDNA sequences; these dogs have profiles that are shared with only 1 to 5 per cent of other dogs. A further quarter have sequences that are likely to be shared by fewer than 1 per cent of dogs, making mtDNA analysis more effective. 鈥淚t鈥檚 kind of a roll of the dice,鈥� says Joy Halverson, director of , an animal DNA testing company in Davis, California.

Occasionally forensic scientists find a particularly rare mtDNA profile that can make or break a case. For example, canine mtDNA helped to convict a burglar in the US who broke into a home and put the owners鈥� two barking pekinese dogs into a lit oven. The larger dog, Dexter, managed to escape, leaving a droplet of blood on the defendant鈥檚 shorts in the process. Elizabeth Wictum, who runs a at the University of California, Davis, worked on the case and calculated that the likelihood of the mtDNA from this droplet belonging to another dog was 1 in 100. 鈥淭he actual DNA sequence was only seen in dog breeds originating in Asia 鈥� and because they were pekinese, that gave further evidence beyond statistics,鈥� she says.

So far there have been a few dozen instances worldwide where evidence from dog DNA has been presented in court or used to put a criminal case together. Advocates of the approach know that if it is to become mainstream, they must prove its worth. 鈥淚t is the same technology, but has not had the same scrutiny that human DNA evidence had to go through,鈥� says Wictum. She and her fellow dog DNA researchers aim to change this by mirroring the processes that human forensic DNA technology has been through 鈥� validating everything, publishing in peer-reviewed journals and being transparent. The rigour is starting to pay off: in July the American Society of Crime Laboratory Directors 鈥� the main body in the US assessing human forensics laboratories 鈥� formally recognised Wictum鈥檚 lab, making it the first animal crime lab to gain .

Validation is not the only issue, though. If dog DNA is to become a regular part of forensics, with commercial test kits available to every police force or investigator who wants it, the field needs to be standardised 鈥� as it is with human DNA profiling. At the moment, different dog DNA labs use different sets of marker sites. Kanthaswamy鈥檚 database relies on 18 markers and one sex-linked marker. These were selected from a set of candidate markers identified by an international panel as those with the best physical and chemical characteristics for forensic purposes. The UK database also uses sites derived from this set. Meanwhile, Wictum鈥檚 lab has updated its markers following the publication of the dog genome in 2005, mining thousands of markers over four years to come up with its own set of 15 markers and one sex-linked marker most likely to give clear results.

Despite these teething problems, use of dog DNA in forensics is steadily growing. Both Wictum and Ogden say that in recent months they have received more enquiries from investigators looking at reopening 鈥� cold cases鈥� 鈥� those long closed owing to a lack of evidence. And in August, for the first time, Wictum and colleagues attended the meeting in Reno, Nevada, to tell investigators about the possibilities of animal forensics.

鈥淐riminals are getting smarter and smarter,鈥� says Wictum. 鈥淭hey read the same journals and watch the same TV shows 鈥� they know to wear gloves. But it鈥檚 still possible to transfer dog hairs in the sheet wrapped around the victim.鈥� Halverson makes a similar point. 鈥淭he proximity of domestic animals in our lives is so huge,鈥� she says. 鈥淒og hair just gets everywhere鈥︹��

鈥淭he proximity of domestic animals in our lives is huge. Dog hair just gets everywhere鈥�