Why does it take so long to get DNA results?

Factual Questions
1

Obviously this came to my mind with the reported deaths of the Hussien boys but I’ve often wondered about this. When a body is found in a high profile case, it typically takes a week or two to get identification results for DNA analysis. What is the process for doing this and why does it take as long as it does to get the results? I can understand if it’s a less visible case and there is a backlog at the lab but I assume that the higher profile cases would get pushed to the front of the line.

Haj

2

They have to have time to tinker with the results :wink:

3

That’s actually a good question. I work in a genetics lab, and the longest test we do (Southern blotting) takes about 4 days from getting the sample in to releasing results.

Now, we don’t do anything forensic, but if I had to guess, here are some possible factors that would add to the time:

  • Extraction difficulties. It’s a lot harder to get useable DNA from a smear on a piece of charred rubble than from a nice tube of blood. This could even lead to them having to repeat tests if they get questionable results.

  • Redundancy. I imagine with big cases like this they’d have multiple labs do the test multiple times, and probably have multiple people look at and interpret the results, all of which takes time to coordinate.

  • Transportation. Even in my lab, it takes a couple of days, on average, for a sample to get from the doctor’s office to our lab, and almost all of our clients are in the US.

4

What is the shortest test? The US says it’s already verified Sadam’s sons bodies by genetic testing.

5

Are you sure? Do you have a link? The last I heard was that someone who knew them said it was they.

6

No. They have identified them by sight. Their was enough left of their faces for people they know to confirm their identity. Odai was further identified by the scars from when he was attacked a few years ago.

Thanks, Smeghead. That’s probably very close to the answer to the GQ.

Haj

7

Maybe. I misread the first article I came accross, but while his MSNBC article does say that preliminary ID was done by physical appearance, DNA samples were taken and that with state of the art field testing gear, it’s possible to complete an ID within a day.

I have difficulty believing they would make such certain announcements without something more than comparing bodies to some photographs.

8

They got people who actually knew them to identify the bodies.

Haj

9

If that’s the case, I would suspect the “field” tests will give you a preliminary result that would be confirmed later with more extensive testing.

10

Most of the tests are PCR or RFLP (restriction fragment length polymorphism) tests. The delay probably comes in transporting a sample, possibly while refrigerated, to a place where these can be done cleanly. Another long delay would be in acquiring reference samples, but I assume the government has already identified those, either from hair samples from their palaces or relatives.

PCR is simple and easy, but it is not the standard for identification. It also is very prone to contamination and is totally uninformative in many cases.

RFLP analysis is harder and requires more DNA and takes longer but is not prone to contamination and can be more informative than PCR. It is Southern blotting based, as Smeghead mentioned above.

A brief rundown of the procedures. For PCR: DNA is very crudely extracted (under 30 minutes probably), and this is often optional depending on the exact PCR. One makes a mixture of enzymes, primers, DNA, and assorted buffers and runs it in a thermocycler for usually 2-6 hours. It is then run on a gel for another few hours to overnight. This is the most rapid.

For RFLP, DNA is finely extracted (6 hours or more of protocol). It is then digested with several restriction enzymes (2 hours to overnight). It is then run on a gel slowly (usually overnight). The gel is then washed and denatured for 3 hours or so, and then transferred onto a nylon or nitrocellulose membrane for at least 4 hours (also can go overnight). The next morning, the membrane is washed well. Concurrently, you prepare a radioactive probe by labeling another piece of DNA. This is hybridized, usually overnight. The next day it is washed well and set up with radioactive film. Depending on the strength of the probe (length and sequence and freshness of isotope used for labeling), this incubation lasts anywhere from 30 minutes to 2 weeks.

Of course this process can be streamlined, but the above is how I do it. Count on 4 days, at least, as Smeghead said. First day extract DNA, cut DNA, and run the gel overnight. Second day, wash the gel and transfer it. Third day, label probe, hybridize with the membrane. Fourth day, wash and set it up with film.

11

To clarify. PCR is polymerase chain reaction, by which a piece of DNA is exponentially amplified by heat stable polymerases. This can only be used to look at parts of the genome where there is some sequence length changes between control and test. This is true of certain simple repeats, for instance STRs (short tandem repeats).

RFLP uses restriction enzymes which cut (usually) at 6 base pair inverted repeats. Since this queries the direct sequence of DNA, there are many other kinds of DNA polymorphisms (variations between individuals) that can be used. Many RFLPs or STRs are necessary in order to create a DNA ‘fingerprint’ to get near certainty.

12

<off topic>

You use a 6 hour extraction process for Southerns? Ours only takes about 40 minutes, hands on time, plus maybe another hour to resuspend it. Of course we’re using genomic DNA, which might be easier.

We also use air cyclers that can do a 30 cycle PCR reaction in about 40 minutes. :smiley:

</off topic>

13

Yeah we still use heat pump thermocyclers for the PCR or RoboCyclers. They aren’t that fast, but they get the job done. (really off-topic) Yesterday I ran a 36 hour protocol and amplified an 18 kb product. That shatters my old personal best of 4 kb.

For flies, the clean prep protocol we use involves homogenization, then lysis (1 hour), then isopropanol precipitation and wash (30 minutes), then RNase A treatment (1 hour), then phenol chloroform extraction times 2 (30 minutes), then ethanol precipitation and washing and resuspension. (30 minutes). So that’s about 3.5 hours, if you are rushing. I’m sure there is a faster way, but we get nice clean DNA and it works. Then you have to factor in test cuts and gels and spectrophotometry just to make sure it’s nice and clean.

I’m sure there is a Trizol or a quick-and-dirty cesium chloride prep that is shorter. I just know what I do, though, and I know what I do works. You know, that scientist-and-his-protocol superstition that we all carry around.

Probably those in the bidness of speed, like the cops and the military, know how to do this fast. I wouldn’t put it past them to use some crazy techniques like real-time PCR with fluorophores or HPLC polymorphism analysis after PCR, which would speed everything up and increase resolving power.

14

So where did the US Govt get these “reference samples” for Hussein and his sons?

Have we captured Hussein’s wives, and taken blood samples from all of them? And don’t you need samples from both parent to get absolutely positive identification?

And I’d think it would be unreliable to count on samples from their palaces – too easy to accidentally get a hair sample from one of his aides, or the guy who does the laundry, or the one who makes up the bed, etc.

15

The papers today said that they used a presidential aide and also dental records. I have no idea where they acquired the dental records.

The easiest way to get reference samples is either to sample mitochondrial DNA from someone in the direct maternal chain (their mother, their maternal aunts and uncles, their maternal first cousins providing an aunt was their mother, their maternal grandmother, etc.) or get Y chromosome DNA from someone in the paternal chain (their father which is unlikely, their paternal grandfather, paternal uncles, paternal male first cousins providing the father was an uncle, etc.) These wouldn’t by the absolute best way of doing identification, but they would be the easiest.

I believe Saddam Hussein has many sisters and brothers and other members of his clan for which one could get better samples than Y or mtDNA. These would be autosomal polymorphisms, but in order to make any sense of these, you need a bunch of different relatives and a very definite pedigree. The best way of course would be in identifying samples that are known to be of the brothers or the father. Yeah, they may have identified 100 different people from the palaces, but that could be sorted through. The chance of identifying the same piece of DNA from a few presidential palaces and one unidentified body must still be quite small, and therefore adequate for genetic identification.

16

hajario, as you can see above, it doesn’t take all that long. Another reason is because after they get the results they wait to annouce them. They know the results but wait before informing the public. There can be several reasons for this.