Is it possible to tell human and simian DNA apart?

Hawk replies:

Not likely. Ignorance being what it is, these days we rarely get out of the bunker.

It’s little wonder that DNA analysis is confusing. Given a tissue sample of unknown origin, it’s possible to use DNA to identify the species. It’s also possible to determine the likelihood that two DNA samples come from the same individual–that’s what is commonly meant by “DNA fingerprinting.” But for reasons we’ll explore, it’s generally not practical to do both things at the same time.

To better understand the difference between individual- vs. species-level analysis, it’s easiest to look at the past. Prior to the introduction of forensic DNA analysis in the last 20 years or so, species identification of forensic samples was performed using an antiserum. In this method, host animals were inoculated with purified protein from a foreign animal species, which produced an immune response in the host animal (e.g., the host animal produced antibodies to the purified protein). These antibodies, which are “custom-built” to recognize and bind to the purified protein, could then be isolated and used as an analytical reagent for identifying that particular species.

But they provided only a yes-no answer. For example, if a sample of blood tested against anti-human antiserum reacted positively, that told you that the sample contained human proteins. (The truth is that the anti-human antiserum also reacts with many primates, but more on that later.) If the sample didn’t react, though, all you knew was it contained no primate contribution. That’s not an identification; you only knew what the sample wasn’t. You had to test an unknown sample against an entire library of antisera in order to get useful information.

That library was by no means complete. Even today, a full forensic antisera library will include human, cow, chicken, deer, pig, horse, sheep, rabbit, goat, mouse, rat, dog, and cat. In the past, antisera have come and gone (the host animal doesn’t live forever, you know), like pigeon, bear, and alligator. If the sample were from a species not commonly included in the antiserum library–say, an elephant–you’d never identify it.

A similar situation prevails today. Even if you’re using a modern DNA analysis kit, the kit still has to include DNA sequences unique to a species in order for you to identify that species. In other words, you wouldn’t know that a sample was from an elephant unless you suspected you were looking for an elephant and had the right DNA test. If the sample were from an aardvark, all your tests would come up negative.

Now let’s turn to DNA “fingerprinting.” Modern forensic DNA analysis isn’t specific to humans; it uses special DNA sequences called “primers” that can also be found in just about all upper primates. In fact, if a forensic biologist weren’t told at the outset about the possibility of a non-human DNA source, he or she would most likely make an error.

Why does forensic science allow what most would consider an inaccuracy in DNA testing? Quite simply, because it’s rare that a non-human primate–say, a silverback lowland gorilla– is suspected of committing a crime. Even in the odd case, such as an attack by dogs, wolves, coyotes, etc., crime scenes involving animals in general are quite telltale, with a lot of blood and hair. Or in the case of primates, banana peels.

SDSTAFF Doug adds:

It might be worth noting that if non-human DNA is involved, there is a massive global archival resource called GenBank which stores sequence data for nearly every organism that has ever been sequenced; this gives a reasonably accessible standard of comparison, with fair odds of turning up a match for many of the most common plants and animals that have had sequencing done. I’m sure this includes many non-human primate sequences, and could thus potentially tell you which species of higher primate a given DNA sample came from.

SDSTAFF Hawk replies:

You’re suggesting that the DNA code be “read” and then compared against GenBank. While that’s possible, it’s impractical–the Human Genome Project took dozens of labs almost a decade to read the three trillion base pairs in human DNA. If I’m not mistaken, most of GenBank’s sequences are of unicellular organisms; of the multicellular organisms, the library is far from complete. Still, this method is an example of “forward” analysis (i.e., when the source is unknown).

Could you just read the unique part of the DNA sequence that says it came from an aardvark, platypus, or Billy Carter? Not really. The genetic code comprises only four bases. It’s the particular sequence of bases that separates us from Billy Carter (and some of us are farther than others).

Is it possible to identify only those segments of DNA that are specific to a family/genus/species? Sure. All you’d need is to identify a segment of DNA unique to all members of a given family/genus/species, and then create a specially-labelled complementary segment. If the segment (what we call a “DNA probe”) finds its complementary sequence in the unknown sample, then you’d know that you had that particular family, genus, or species. If the complementary sequence is not present, then there would be no binding (and no indication as to the species of the source). This is an example of “reverse” analysis (i.e., when the source is suspected).

Could you have a magical cocktail with DNA sequences to all known species? Yes, but that’s like saying you have a magical key ring that contains every single possible key combination–it’s just not practical.

I had a case last year where a man received a package in the mail that appeared to be leaking blood. He notified the police and they found that the package contained a heart. They asked me to test the heart to determine the species. Since I was going to use the antibody test, I asked the detective, “Do you want to know if the heart is human? Or just what species the heart is?” If the heart were human, there certainly was a problem, but if it wasn’t, the species might not matter–merely knowing that it was a non-human animal heart might be sufficient grounds for charging the culprit with criminal mischief.

On the other hand, we also see cases from wildlife agencies involving possible violations of hunting laws. In order to prosecute a man for killing a deer out of season, you first have to prove that the chunks of meat he’s carrying are in fact deer. In cases like that, the standard forensic antisera library is fine: deer, cow, pig, chicken, horse, sheep, and human. If it’s one of the others (rabbit, mouse, rat, chicken, dog, or cat), who cares? Those species aren’t hunting violations.

What if it were a different law, like cruelty to animals? The same questions come up: Do you need to know what animal it is (or was) to prove cruelty? Cruelty is cruelty (presumably you can tell that it was an animal, as opposed to a vegetable or mineral). If the animal were human, you’d be looking at assault or murder charges, but again, it’s a yes-no sort of thing: is it human, or not? If not, does it matter? My point is, in doing DNA or tissue analysis, you need to know what you’re trying to find out before you start.

SDSTAFF Karen asks:

What is Billy up to these days?

Hawk

Send questions to Cecil via cecil@straightdope.com.

STAFF REPORTS ARE WRITTEN BY THE STRAIGHT DOPE SCIENCE ADVISORY BOARD, CECIL’S ONLINE AUXILIARY. THOUGH THE SDSAB DOES ITS BEST, THESE COLUMNS ARE EDITED BY ED ZOTTI, NOT CECIL, SO ACCURACYWISE YOU’D BETTER KEEP YOUR FINGERS CROSSED.