DNA Fingerprinting

I am so proud that my comments resulted in this column http://www.straightdope.com/mailbag/mdnaanalysis.html

Perhaps it is just my prejudice, but it seems that you agree with me. If you had a lineup for a crime that included me, an orangutan, a gorilla, a chimp, and a nun you could run a DNA fingerprint and determine which of us had done the crime. Or if you had a sample of blood and were asked to determine which of the above species it had come from you could do that, although it would take a lot of work. However, for practical purposes you could not take a DNA sample from a crime scene and say not only did I not do it, but it wasn’t even done by a human. So while speciation of DNA is theoretically possible in reality it is not and could not be done by normal forensic methods.

Can’t you narrow down the species by the number of pairs of chromosomes? If, say, there are 19 pairs you know it isn’t human, and it has to be something that has 19 pairs (I have no idea what if any animal class/order/family has 19 pairs)


Yes you could. For instance, chimps have 23 pairs of chromosomes vs. 24 for humans. However, looking at the chromosomes, known as karyotyping, is not part of the DNA fingerprinting protocol.

Billy’s pushin’ up daisies, since several years now. His DNA, always fairly far from normal, is surely further deteriorated to be not much different from Dubya or a proboscis monkey.

Actually, I’m working on a (grad school) project on this topic. We’ve got some DNA, and we need to find out if it belongs to a chimp or a gorilla. To do this, we use PCR to amplify a transposable Line 1 element. Chimps have it, gorillas don’t, humans have a duplication, so it shows up as two bands. It’s easy and inexpensive. You can use sequences, because GenBank has LOTS of primate information. However, you need to use a commonly sequenced gene, such as those in the mitochondria. Also, sequencing can be expensive; we try to go with the cheaper option when possible.

I’m curious: what species was the heart in the box (if you can tell us)?

Neat report.

At least in VA, you can be convicted of killing rabbits out of season (they’re covered by the Dept. of Game and Inland Fisheries hunting regulations and can only be taken in season). So, while there may not be a mouse, cat, dog, or human season, there is a wabbit season (Fire!)

Sorry, but you got that reversed. Chimps have 24 pairs, humans have 23.

Not the case. Fortunately for DNA testing, degradation doesn’t change the genetic sequence. It will definitely cause fragmentation of DNA strands, but if the DNA fingerprinting you perform doesn’t require long stretches of DNA, (e.g. what are known as short tandem repeats), then you can get a full DNA profile of Billy’s daisy spawning corpse.

Since the primers used in a standard DNA profile are human-specific, we can definitely tell he’s not a monkey. And we can even tell he’s not Dubya, assuming we can get a sample from either guy to compare it to.

I think there’s confusion by this bit about not being able to test both the species and the DNA profile. While you can’t run the same sample through both tests, if you have a large enough sample you can split it into two samples, and then run one through each process. No?

So what constitutes a large enough sample to split into two? Isn’t there a process for culturing a DNA sample to increase its size?

Yea, there’s PCR, or polymerase chain reaction. What it does? Simply replicate the DNA many many times using some nice enzymes (found in viruses and/or bacteria).

A DNA profile is designed to differentiate between individuals of a species. Mostly this is done with humans, although various aspects of forensics and various other fields have produced the ability to get a DNA profile of a specific individual animal. There is a cat DNA profiling kit available called MeowPlex. And IIRC, since horse breeding is big business, there are profiling kits for horses too.

Now you technically can use these DNA profiling kits to determine species. A sample from a cat would yield no DNA using the human DNA profiling kit. A human sample would give a profile.

But practically speaking, it would be wasteful to do this. You’d be using a lot of expensive resources to get a lot of negative results. You’d be better off doing antiserum tests on a sample to determine the species. Once you know that, you can then get a profile for that species, assuming there is a DNA profiling kit available for that species.

This is the usual procedure in our business. It’s rare, but when you have a murder that takes place across the street from an abbetoir and the suspect claims the blood on his clothes is from tripping over a decaptated cow, it’s very useful.

From the column: “…whether DNA fingerprinting can differentiate between human and simian DNA”

From the thread that started this whole topic: “Given sample A and sample B of DNA, can a it be determined with a scientific certainty that, say, sample A is human and sample B is simian?”

There is a subtle yet important difference between these two questions.

DNA fingerprinting will not tell you the source of the DNA, it will only tell you if one sample comes from the same source as another. This is because DNA fingerprinting is concerned with the size of DNA fragments, not the sequence.

However, other techniques can be used to determine the source of DNA. GenBank is a much greater resource than SDSTAFF Hawk realizes. If something has a nucleic acid sequence, whether it’s a plant, animal, bacteria, virus, or manmade sequence, Genbank has at least a partial sequence of its genome. And in virtually all cases (including ours) it’s not necessary to have the entire genome. Of the 3 billion or so base pairs, only a couple thousand base pairs unique to either man or ape would be sufficient. One can use GenBank to find these unique sequences, then use them to probe your unknown DNA samples.

“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.”–SDSTAFF Hawk.

Hawk underestimates recent advances in genomics. The magic keyring is called a DNA chip. A single chip may not have every combination, but it can hold several thousand different fragments-- so a few chips should be able to screen a huge number of species. Species-identifying chips are already used to detect pathogens, and it’s only a matter of time until the “physical sequence database” required for DNA chips catches up to the “paper sequence database” of GenBank.

Hawk is remarkably ignorant of how GenBank is used. It is used ROUTINELY in EXACTLY the way he claims is impossible. One does not even need chip technology to do a homology search. NCBI has a program called “blast”, which can compare any DNA sequence you enter with the entire database or just portions of it. This can take as little as a minute or as long as an hour, depending upon sequence length and load.

I use blast on a daily basis. Indeed, if you give me a sequence of unknown origin, within less than a day I could tell you how many, if any, sequences in GenBank match it, how closely they match it, etc. If it has homology with a human, rat, mouse, or other GenBank genomic sequence, I could tell you exactly where in which chromosomes this appears. I could also tell you the identified homologues from other species. This is entirely routine work these days–no big deal and no insurmountable problem. Hawk needs to go re-take sophomore biology.

Dogface, you’re absolutely correct in everything you said. If you had a sequence of unknown origin, a blast search would give you an answer without much effort.

However, getting that sequence from a sample of unknown genomic DNA would be incredibly difficult. Given the high homology of DNA among species in general, let alone among primates, you’d probably have to do a lot of sequencing before you found a region unique to an organism. Hawk wasn’t that incorrect in thinking that you’d have to sequence the entire genome.

That’s why I suggested a chip may be the best way to go. Getting such a chip made today would be difficult-- I’m sure the database of physical DNA specimens is nowhere near GenBank’s sequence database, but in the near future, such a procedure will be almost as routine as a blast search.

Let’s deal with this series of issues one at a time. First of all, I’ll address the issue of my credentials. I graduated in 1991 from the University of Southwestern Louisiana with a B.S. in Pre-Medical Sciences. I have been working in the field of forensic science since that time, specializing in Biology (there are numerous other disciplines, including Chemistry, Trace, Document Examination, Toxicology, and Arson, to name a few). In my spare time, I enjoy sharing my knowledge with others by answering questions in this forum. While I am not versed in all areas of biology as much as I’d like, I do live in the shadow realm of forensic science, and it is a place with certain considerations that many people don’t realize. Frequently, these people are in academia or research, whose studies are often conducted in controlled, sterile conditions. In doing this particular kind of writing, small detours are taken, not to confuse the reader, but avoid lengthy articles on issues that are tangentially related to the original question. My following comments use capitals to highlight key ideas, not to exclaim. The following, then, are my views and mine alone, and do not reflect those of the Straight Dope:

Please take into consideration that a forensic sample can have a number of problems with it, problems that academicians and researchers can not appreciate. In the worst case scenario, a forensic sample is (a) small, like the size of a period (.), (b) degraded, like from an article of clothing taken from a body that has been exposed to the elements, © contaminated, like from maggots, and (d) aged, like from a body that has been exposed to the elements for an unknown length of time. Keep in mind that any given sample that comes through my doors can have varying degrees of all four factors. Unlike research, I do not get to ask for more sample or for better sample; I simply get what I get and that’s it. On the chromosome issue, I ask this: “How certain can you be that you had all of the chromosomes in the organism’s genome? Could any have been lost due to sample size or decomposition? And how many other organisms have that many chromosomes?” This method could certainly not identify the species, to the exclusion of all other species on earth.

The heart turned out to contain porcine (pig) proteins. Thus, the animal it came from is either a pig or is at least related to pigs. By “related to pigs,” that could include wild boar or maybe even warthogs, for all I know (I’ve never been inclined to check how closely related such animals are, because that particular question has never arisen within my experience).

It is of interest to me to note that Virginia has a rabbit season, for it goes to the very heart of forensic work: What if the sample were a jackrabbit? Can you punish someone for killing jackrabbits in the state of Virginia? While I may be splitting hares (sorry, couldn’t resist), the issue is that THE LAW determines what shall be significant.

Having laid down all that foundational material now, let’s deal with GenBank, which is largely a research resource. Why? Because GenBank is not complete. For the standards required in law, it would have to be known, with at least the best scientific certainty at the time, that the DNA sequence comes from that species, and only that species, to the exclusion of all other species on the planet. (I use the term “species” here, but, in fact, we would mostly be talking about the biological order or family of the species).

 What if the sequence you obtain (which could even be hundreds of thousands of base pairs long, none of which may not contain unique sequences) isn’t in GenBank?  What would you do?  What if the DNA is degraded, giving you only fragments that are less than 300 base pairs long?  What would you do?  (Remember, you can’t ask for more or for better samples; this is a real life situation.)  Does the sequence contain those species-specific sequences, or does the sequence contain individual-specific sequences?  My point here is that in forensic science, WE JUST DON’T KNOW WITH ENOUGH CERTAINTY TO TESTIFY IN COURT TO THOSE FINDINGS, especially if someone is going to be fined, imprisoned, or executed because of our analysis and testimony.  Furthermore, I believe that the process of using “GenBank to find these unique sequences, then use them to probe your unknown DNA samples” is an extreme oversimplification that ignores questions like, “What if it’s not human?”  “What if my sample is human, but the DNA is degraded at that particular site?””What if my sample is human, but is so small that it failed to yield enough DNA to analyze?”  I’d get a negative result and I’d issue a report that the sample isn’t human in origin… and I’d be wrong.  That may seem trivial, but consider the weight of my job:  when I’m wrong, I either send an innocent man to jail or death, or I’d set a guilty man free into society.
 This ain’t no “CSI:  Crime Scene Investigation.”
 This ain’t no classroom.
 This is real life.
 Chip technology is useful only in that it has to be “programmed” WITH KNOWN SEQUENCES.  Those sequences would have to come from ALL SPECIES ON EARTH in order for you to use chip technology to identify the species.  (While we’ve been dealing strictly with animal species, there is a good point that the net could be cast wider to include viral, bacterial, algal, and botanical sequences as well, and now the task becomes virtually impossible.)  Even if you didn’t address all species, but let’s say a significant majority of them, you’d still be faced a self-admitted gap in your knowledge.  I will admit that while chip technology could be feasible someday, it isn’t PRACTICAL, because the issue of “What species is it?” doesn’t come up often enough to justify manufacturing the chip in the first place.  If you don’t believe me, take this as a lightning strike of Inspiration, take this idea to market, and see what happens.
 My “remarkable ignorance” notwithstanding, even this excerpt contains the very holes with which I have been attempting to fill with enlightenment.  I’m not as wrong as you think I am, Dogface, nor are you as right as you think you are; perhaps your name-calling should be reserved to those more deserving of it, but as far as I’m concerned, it is unfitting behavior for an adult.  Anyway, if I gave you a sequence of unknown origin, you could use GenBank to tell me one of three possibilities:

(a) the sequence can be found in species X, according to GenBank;
Yes, that’s fine. Next question: “Isn’t it at least possible that this sequence belongs to yet some other species, a species that hasn’t been discovered yet or even, in known species, genomic sections that haven’t been analyzed yet and entered into GenBank?”
(b) the sequence can be found in species X, Y, and Z;
If the sequence is to be found in more than one species, that’s not SPECIES-SPECIFIC, is it? That is the original question. Next question: “Isn’t it at least possible that this sequence belongs to yet some other species, a species that hasn’t been discovered yet or even, in known species, genomic sections that haven’t been analyzed yet and entered into GenBank?”
© the sequence is not listed in GenBank.
If the sequence is not to be found in GenBank, then we are back where we started, except that we have a sequence that is uninformative and have irretrievably lost a portion of the sample (if not the entirety of the sample) to get it. And now there’s a guilty man free in society.

At the risk of fanning a flame war: Hawk, you need to be less defensive. Dogface’s comments may have been incendiary, but given your tone and gross inaccuracies in the column, you were asking for it.

Your critical error is that you’re forcing the question to fit your background. This is a scientific question, not a forensic one. But you’re looking at it through forensic blinders.

I think you’ll agree that this topic isn’t 100% up your alley. That’s fine; perhaps there are no other SD staff with a stronger molecular biology background. However, there are readers such as myself with a stronger background, and we’ve all said the same thing: you can find out whether a given DNA sample is human or ape by using GenBank. Your mistrust of this database is baffling, but if you really want me to, I will do a blast search and tell you what probes I’d use for this experiment.

Another complication you assume is that the source of the DNA is utterly unknown and could be any species, even ones not yet discovered (in which case you’d be screwed no matter what technique you used). The question is simian vs human, but the wording of the question is ambiguous, so I don’t fault you. However, a more plausible inference is that the reader heard somewhere that human and ape (specifically the chimpanzee) genomes are almost identical, and was wondering if the two could be distinguished.

If the unknown DNA could be from any species, you’re left with two options. The first option is to sequence the unknown DNA and do a blast search. If this fails, then any technique would fail. No one claims that GenBank has the complete sequences of the genomes of every organism, but if GenBank does not have it, no other database would. And by definition, to identify the species, you need to compare the unknown DNA to something.

The second option is to work in reverse-- instead of taking the unknown sequence to GenBank, take GenBank to the unknown sequence. This is done using a DNA chip. Again, Hawk underestimates chip technology. He’s correct in believing that species identification via DNA chips would be impractical, that there’s no way I would be able to buy such a chip “off the shelf.” What he fails to realize is that most chips are tailor-made, i.e. you tell the manufacturers what you want on the chip. It is irrelevant how many other people want the same chip I do, what kind of “market” there is. If I wanted a chip with my DNA on it, I’d have it in a couple days. The only problem with constructing a species identifying chip is the availibility of DNA from all the species, but as I’ve said before, that’s growing.

Given this incredibly hypothetical situation, this is how to identify unknown DNA. Sequencing would be thorough yet slow. Using chips would probably get you the answer a lot faster, but screening all possible species isn’t feasible.

This is straying quite far from the original question though, which was about distinguishing human DNA from that of apes. Again-- it isn’t difficult to find, via GenBank, sequences that are uniquely human, and to make a probe to those sequences to test your unknowns.

<< This is a scientific question, not a forensic one. But you’re looking at it through forensic blinders. >>

Sorry, Noose, but I disagree with you. The person asking the question didn’t start with “In a science theoretic sense, is it possible…” And the link says, “…can a it be determined with a scientific certainty.”

SDStaff Hawk took a forensic approach to the question, and a very practical one. He deals with the question of identifying DNA bits in a practical, day-to-day sense, and that’s the approach he took in responding. He didn’t reply in terms of the mathematical combinations of DNA or probabilistic results; he didn’t reply in terms of molecular biology or philosophy, either.

Noose, if you would like to write a molecular biologic response to the question, please do so. Email me, and I’ll see about getting it included as an addendum to the Staff Report. If anyone would care to write a probabilistic response (how many “hits” is needed to identify a strand of DNA as coming from a human, with what probability of certainty?), ditto.

The question, as asked, is a hypothetical one, not practical. Taking a practical forensic approach isn’t the best way to answer this question, because practically speaking, one would not even use a DNA test to determine species. In a forensic setting, I imagine it’s virtually impossible to find naked DNA-- it would likely still be in blood or tissue. As Hawk discussed in his column, one would do antibody testing on the blood or tissue to determine the species.

Given the impractical constraint that we must use DNA to determine species, I think it’s reasonable to assume such a hypothetical exercise would be done in a molecular laboratory, where DNA quality, quantity, and purity are less of an issue.

As for a response to the column from a molecular biological perspective, feel free to use the relevant parts of my first post in this thread. If you need me to go into more detail, let me know.

Well, Noose, I guess we’ll agree to disagree. On a scientific theoretic basis, OF COURSE simian and human DNA are different, they come from different species. So, the only interesting question is whether on a practical basis, we can distinguish them.

And I was way too lax before, we’re not going to have any flaming in this forum. I apologize, I was busy and didn’t notice the tone of Dogface’s comments. So, let me be very explicit:

Dogface, calling a member of the Straight Dope Science Advisory Board “ignorant” or making a comment about “re-taking sophomore biology” is entirely out of place in this forum. You want to do that, you go the forum called BBQ Pit.

In this forum, you may challenge the statements, but not the person. If you are unaware of the difference, you should perhaps re-take a course in good manners. :slight_smile: