Talk:DNA profiling

Partial Matches (and UK Statistics)

There should be a section (or an article) on partial DNA matches, their usefulness to law enforcement, and the controversies going on now. Also, the FBI's new interstate policies should be noted. Should this be a section or its own article?

---

Separately; is that "7 million in the UK database" stat correct? I heard an interview with Denver's DA Mitch Morrisey in which he cited the number at just over 3 million. Anyone have a source for this?

---

The US combined national database has about 5 million. The UK database is over 4 million. CharlesBrenner 27 Jan 2007 (UTC)

---

As of May 2007, the FBI's NDIS webpage states that there are 4,582,516 Total Convicted Offender profiles in the system. [1]. From the minutes of the House of Commons, "Joan Ryan: As of 10 June 2007, there were an estimated 3,976,090 individuals on the National DNA Database (NDNAD)." [2] Furthermore, they estimate that 13.7% of their profiles are duplicates [3] This handily makes the US's NDIS the largest DNA databank in the world. Sekiyu 04:55, 16 August 2007 (UTC)

2003

I don't understand how a 1 in 5 million chance translates into 30 people who would match the profile from a population of 60 million people. Wouldn't that be 12 people (and not 12 other people, 12 people in total)?

Marisano

---

Agreed. I changed 30 to 12.

Zashaw 02:06, 18 Aug 2003 (UTC)

---

This page seems redundant with DNA testing -- the articles should probably be merged, or, if there's a distinction between the articles, it should be made explicit.

Zashaw 04:24, 18 Oct 2003 (UTC)

So what are the advantages of this system

---

The second paragraph mentions PCR, the section on the DNA fingerprinting process doesn't. Which one is correct? Furthermore, what kinds of restriction enzymes/PCR primers are being used? AxelBoldt 16:21, 29 Jul 2004 (UTC)

---

The article initially states "...STR profiling...is distinct from DNA fingerprinting...", but later categorizes STR analysis as a type of DNA fingerprinting. Is something missing, or is STR profiling distinct from STR analysis? Ian Glenn 21:34, 3 April 2007 (UTC)

No, they are the same thing. It is a kind of DNA analysis using PCR, which uses primers. RFLP, distinct from PCR, uses restriction enzymes. RFLP is little used now.165.176.123.2 (talk) 19:54, 28 February 2008 (UTC)

Identifying a deceased person

Can someone explain in the article how a deceased person can be identified by genetic fingerprinting?

I presume DNA is taken from a sample of a body part. Is it compared with a sample known to come from the putative person, before they died? Or is a comparison made against samples from blood relatives? If the latter, what are the principles involved? Is the situation the same as in a paternity test? Thincat 14:53, 24 Jan 2005 (UTC)

It usually is compared against their parents / dental records / hospital samples and things like that. The procedure isn't fixed: the onus is on the law enforcement officers to use the sample to prove it in a number of ways. -- Natalinasmpf 02:10, 31 Mar 2005 (UTC)

I UNDERSTAND COMPLETELY.

Identifying ancient, deceased individuals

I have heard of extracting DNA from the fingerprints of the Ancient Egyptian laborers responsible for building the tombs. This is because fats and oils (such as sebum), along with salts and sweat, are exuded from the fingers and hands, and deposited on those objects they happened to touch. This happened thousands of years ago, back when the great pyramids were first built. The fingerprints were sealed up, and left untouched for thousands of years.

Does anybody have more information on this?

The fats and oils usually deposited in fingerprints are naturally subject to degradation with time, and because DNA is a protein, usually wanting in the oils exuded by the skin, that rules out DNA fingerprinting. But since the Ancient Egyptian laborers had to manage large blocks of stone by cutting, sliding, and rolling, some of those fingerprints may have been bloody, and that was where the researchers managed to get their DNA.

DNA is not a protein. It is a nucleic acid. They are completely different categories of biomolecules. Whoever is going around dispensing information about DNA, without even knowing what it is, needs to cut it out.

Proteins are polymers of amino acids, which have a carbon center, carboyl, amino and -H termini, and a disctinctive side chain known generically as "-R." Nucleic acids have a sugar-phosphate backbone, to which are attached nitrogenous bases. 165.176.123.2 (talk) 19:59, 28 February 2008 (UTC)

I agree, and it's jolly annoying that whoever is giving out this false information doesn't even have the manners to sign their posts. Declan Davis (talk) 20:33, 19 September 2008 (UTC)

betch!!! hoe scum bag!! —Preceding unsigned comment added by 70.231.240.67 (talk) 00:50, 9 October 2008 (UTC)

Extracting DNA from fingerprints

When it comes to DNA left at crime scenes, is it routine to extract copies of DNA directly from the fats and oils of fingerprints, especially for the purpose of comparing them to blood samples found there, to determine how many individuals may have been at the scene of a crime?

The fingerprints have to be bloody in order to extract DNA from them. The fats, oils, and salts associated with fingerprints usually lack DNA, a protein.

DNA is not a protein. This is an error I have come across repeatedly on wikipedia discussion pages. DNA is a nucleic acid, which is entirely different from a protein.

If the fingerprints themselves are usable prints (not smudges, for example), they would themselves be used to identify persons from the scene. DNA may get all the glamour nowadays, but latent prints are still very widely used for forensic purposes. In general, if it is thought that a suspect touched something at a crime scene, the item would be processed by latent print examiners.

That said, while the salts, lipids and proteins in latent prints are not generally considered a useful source of DNA, what can be done in some cases is to analyze what is known as "touch DNA," in which an examiner attempts to extract DNA from skin cells left behind when someone touched something, such as the rim of a hat worn, or a steering wheel. This is not "routine," however.

Analysis of touch DNA would be done for the same purpose any other DNA samples--to identify persons present at the scene. To say "especially for the purpose of comparing them to blood samples found there, to determine how many individuals may have been at the scene of a crime" is rather too specific to confirm; in fact it sounds like an episode of CSI. Additionally, if there is plentiful (miniscule amounts are required now) blood (or other common sources such as semen or saliva) available, touch DNA would normally be considered unnecessary.165.176.123.2 (talk) 19:48, 28 February 2008 (UTC)

Fetal DNA sampling

Is anyone able to contribute information regarding the methods of sampling fetal DNA, whether invasive or non-invasively? I wasn't able to find anything particularly definitive after an hour of online searching.

It would also be useful to add information about the legality of such procedures (including whether they can be the subject of a warrant) and any risks to the fetus that may be involved.

• Fetal DNA can be extracted from maternal blood after the 12th week of pregnancy (non-invasive), before week 12 an invasive proceedure like Chorionic villus sampling (CVS) and/or Amniocentesis has to be used. I can't comment on the application of law to force either type of proceedure. --nixie 05:08, 11 July 2005 (UTC)

What is Dr. Schneeberger's first name and where in Canada did the 1992 case of fake DNA evidence take place? Please provide a source. Thank you. Dphen 19:56, 10 February 2006 (UTC)

DNA Databanks

With the recent additions by 88.106.31.222, it seems like this article has started to cover more than the scope of just genetic typing. Most of the information about the FSS's databank could be moved to another article just about DNA databanks. Sekiyu 18:25, 26 April 2006 (UTC)

First case

This website lists Tommie Lee Andrews' conviction as 6 Nov 1987 (a quick google search seems to confirm it). This is the year before Colin Pitchfork's conviction is listed by BBC. Can anyone find some source to unmix this? Ted 06:03, 28 April 2006 (UTC)

• Good question! According to the Gene Technology page you linked, the line "Andrews became the first person in the United States to be convicted of a crime based on DNA evidence" is key. The Pitchfork case was in the UK, and he while was indeed sentenced in 1988, the investigation itself occurred in early 1987 (he was arrested on Sep 19, 1987 and plead guilty to the rapes and murders). An interesting side note is that the first time that genetic fingerprinting was used for law enforcement, it was actually in the exoneration of an innocent man: George Howard in 1986. Sekiyu 20:30, 28 April 2006 (UTC)

If he plead guilty, then he was not the first conviction due to DNA fingerprinting. Maybe a better wording would be "first criminal caught using DNA fingerprinting." The Date would then be 1987 (when he was arrested), which would match with the conviction of Andrews. It might be useful if someone with a law background could verify the "exoneration" (which has a more strict legal definition than popular defintion). I have made a few changes to that section to clean up the language a little bit. Ted 00:22, 1 May 2006 (UTC)

Hey what about the first person to exonerated from death row. His name is Bloodsworth, Kurt (i think). You can find him on the innocence project's website and someone's wrote a book about him??

The Southside Strangler case

According to [4] [5], a Timothy Spencer was the murderer of four females in the famous "southside strangler" case in Virginia. He is known to be the first person jailed and put to death in the United States due to a DNA analysis match in his semen on the four female victims he raped and asphyxiated (or other way around). The case was also mentioned several times in the Forensic Files series, where they also claim that it was the first time ever in the United States where someone was arrested due to circumstantial DNA evidence found in his semen to match on his raped victims. I think that should that be mentioned in the article since that case was also in 1987. Comments anyone? (Me | The Article) 08:01, 13 January 2007 (UTC)

DNA mass surveillance

Is there an article giving examples for DNA mass screening ? -- Simplicius 15:05, 28 July 2006 (UTC)

Considerations when evaluating DNA evidence

The sentence "Because of this, arbitrary ceilings were put on match probabilities used in RFLP analysis than the theoretically computed ones." needs to be cleaned up. "Than" calls for an adjective in the comparative degree - higher ceilings or lower ceilings or stricter limits or whatever.

• I've fixed this unclear part and added a reference link to more info about the ceiling principle. Sekiyu 16:49, 8 February 2007 (UTC)

The claims about error rates and ceilings are confused and incorrect, and amount to a partisan rendering.

• 1. The motivation for the ceiling principle (proposed by the 1992 NRC report) was insufficient population data. It had nothing to do with errors.
• 2. The article by Jonathon Koehler mentioned by the referenced 2002 newspaper article could not have been a reason for moving away from RFLP several years earlier.
• 3. The reasons for switching from RFLP to STR were increased sensitivity, easier interpretation, and being less labor intensive. I don't recall error rates entering the discussion. Indeed, errors in interpretation is a relatively unimportant component of possible errors, and STRs because of greater sensitivity are more susceptible to error by contamination.
• 4. Koehler is a psychologist and not a DNA scientist or even statistician. His research scored an error if a lab reported a type of "4.1" when the correct answer was listed as "4" -- for lack of understanding the notation, that "4.1" was not different from "4", but merely more precise, designating a sub-type. Was the study reported in the 2002 newspaper article a different one? Further, despite his assurance to the contrary, his statistical approach is contrived. CharlesBrenner 15:22, 4 July 2007 (UTC)

I'm not sure were it might fit, but would it be worth tying in a reference to the Phantom of Heilbronn in this section? There is recent evidenice suggesting that the DNA matches tying together those 40 crime scenes in Austria, France, and Germany were caused by contamination from the factory makding the cotton buds for DNA swabbing.Occasional Reader (talk) 18:51, 26 March 2009 (UTC)

Isn't the stastical reasoning in regards to coincidental matches flawed? As I understand it, the point is that the chance (risk) of a coincidental match, between two randomly chosen persons, is in theory 1 in 100 bilion...but in reality it is 1 in 1000 - no? At a glance this looks problematic... It seems that it is assumed that there's a probability of 1 against 1000 not only to pick two monozygotic twins, but that they're each other's sibling. Wouldn't the chances of that happening rather be something like 1 in 500.000? And the practical risk then being something in the same neighbourhood (1:500.000,000005 or something like that)? —Preceding unsigned comment added by 212.242.152.183 (talk) 16:27, 28 March 2010 (UTC) Sorry... It would be 1:499.999,999995 - not 1:500.000,000005 :) That is, If I'm correct... I suck at calculating probabilities, so please correct me if I'm wrong...

Independent assortment

Is there any research that substantiates the assertion that the loci in the genes are Independently Assorted? Though this is a nice assumption in court, if it is not true then STR testing would be substantially less powerful then claimed to be. I would refer to Keith Devlin's columns in MAA Online for the months of September and October for further reading on why it is important to first establish the assortment independence as fact before using the power rule to establish its statistical accuracy. RSimione 20:57, 29 December 2006 (UTC)

• Absolutely. Check out [6] and [7]. JFS is one of the foremost peer-reviewed forensic sciences journals in publication. Sekiyu 16:45, 9 February 2007 (UTC)

• • I read Keith Devlin's article; I found it ironic that he would present some specious arguments himself. His argument sounds strong to a naive reader, but his October article hinges on the Birthday_Paradox. If you plug his numbers into the formulas that are listed in that article, you will find that in a database of 65,493 entries, you will expect to have a 50% chance of a match with only 301 people. If you understand the probabilities involved, I think that you will be disappointed by Devlin as well. Sekiyu 17:44, 9 February 2007 (UTC)

1 in 100 billion?

According to the article, "the theoretical risk of a coincidental match is 1 in 100 billion (100,000,000,000)."

This theoretical risk is almost certainly overoptimistic. See http://www.maa.org/devlin/devlin_09_06.html for more details. This should be noted. 71.102.156.213 00:04, 18 March 2007 (UTC)

• Thanks for you concern, however it should be noted that this subject is covered in the last subject of this discussion page, titled "Independent Assortment".RSimione 06:38, 7 May 2007 (UTC)

Also used for plants?

The lead says this technique is used for "individuals," but I believe it is also used to identify the genes of plants. Thus, the lead should be revised. Badagnani (talk) 22:17, 12 December 2007 (UTC)

Copyright violation/copy and paste

The "Fake DNA" section in this article is exactly the same to a section here: http://www.bio-medicine.org/biology-definition/Genetic_fingerprinting/#Fake_DNA_evidence . May be other copy-pastes. Tarthen Brown (talk) 10:31, 13 May 2008 (UTC)

Found one: In Cases on OJ Simpson - these two are the same more or less. Who copied who? Tarthen Brown (talk) 10:33, 13 May 2008 (UTC)

Chimera (genetics)

DNA is not 100% accurate all the time. Please add information and reference Chimera (genetics) A Chimera has 2 sets of different DNA. Blood could be one and skin could be differant. —Preceding unsigned comment added by 24.255.120.218 (talk) 06:08, 3 August 2008 (UTC)

I guess that DNA is always accurate. It's our means of collection and analysis that aren't always 100% accurate. Declan Davis (talk) 20:28, 19 September 2008 (UTC)

Crime work

Given a sample of DNA, what information can we currently know? I guess that technically we should be able to rebuild a person from their DNA. As it stands, what can we do? Can we determine hair colour, eye colour, ethnicity, etc? Or do we just get a complex code that we try to match with another piece of code, kind of like playing the card game snap? Declan Davis (talk) 20:26, 19 September 2008 (UTC)

Chimera confusion in lead

From the lead:

"These loci are variable enough that two unrelated humans are unlikely to have the same alleles, the only exception being the rare individuals which have two different sets of chimeric genes."

This is not an "exception" to that rule, but an extension of it: a chimera is not more likely to have the same alleles as someone unrelated to the chimera.

I am correcting the lead, accordingly. If you object, please discuss in this section and/or cite reliable references that support the claim.

Thank you. —Danorton (talk) 18:00, 4 October 2008 (UTC)

How long does it take?

The article could use expert input on how long each of the DNA analysis methods takes to get results. There is no mention of it here, and there seems to be much confusion among the public on this issue. I often hear comments from people on the radio or from acquaintances that indicate that they think it's as fast as blood typing and cross-matching (minutes to hours), though from what I recall hearing some years ago, it typically takes several weeks to get results. —Quicksilver^{T @} 00:56, 9 October 2008 (UTC)

The article could use expert editing, but expert input into the article is contrary to Wikipedia policy. (See WP:NOR). Anyone is welcome to research that information from reliable secondary sources and include it here. If you're unsure, post the information here first for feedback. Experts can then judge the reliability of the source. —Danorton (talk) 15:31, 9 October 2008 (UTC)

If you're talking about just doing the lab work, it is about a half day's work with current STR techniques. You'd go from some kind of sample, say a blood stain on a shirt, cut a portion of that and extract the DNA from the cells on the fabric using PCI, total time ~20 mins. Next you'd go to quantification of your extracted DNA -- a simple, rough estimate can be performed with a fluorescent method in about 5 minutes. Next the extracted DNA would be taken for preparation for amplification by PCR, total setup time about 5 minutes as well. A standard amplification takes roughly 3.5 hrs, then preparation/running on a CE instrument for roughly 1 hr. Just like that, you've gone from sample to genotype in about 5 hrs of work. The reason labs take so much longer to get results out is that in the forensic process, steps like accurately documenting evidence for future trial or chain of custody issues are time consuming and do not produce any of the required lab work for generating a profile. After a genotype has been obtained, it still needs to be interpreted by a DNA analyst, then those results need to be reviewed by other analysts and reviewers to make sure that everything was done correctly before any kind of results can be released to law enforcement or the public. Unlike TV shows such as CSI, most crime labs cannot clear their plates and devote 100% of their time and effort to a single case as it comes through the door. So while it's not as fast as simple blood typing, it's not necessarily as slow as you might think. Sekiyu (talk) 07:04, 14 April 2009 (UTC)

Phantom of Heilbronn

There is a interesting case in germany where DNA fingerprinting led to the "construction" of a suspected female serial killer attributed with 40 crimes including several cases of murder. It turned out recently that the DNA traces originally attributed to the suspected killer is that of a woman working in the plant that produced the cotton buds used by the police for sampling the DNA at the crime scene. Maybe anyone can make something out of it. My english is'nt goog enough to edit here and i dont have an account here anyway. 88.128.52.77 (talk) 22:52, 27 March 2009 (UTC)

That story is indeed awesome, however I don't think it really adds to this article in a meaningful way, other than cautioning that negative or substrate controls should be run regularly (this is good scientific practice in general, not just in forensics labs). Sekiyu (talk) 07:22, 14 April 2009 (UTC)

DNA profiling vs Full_genome_sequencing

Whats the diffence between DNA profiling and Full_genome_sequencing ? Is it the same? If so, the 2 pages can be merged 81.246.169.193 (talk) 07:37, 20 April 2009 (UTC)

other names

other names are DNA fingerprint and idiosyncratic minisatellite hypervariation profile. Include in article —Preceding unsigned comment added by 81.246.164.113 (talk) 06:56, 21 April 2009 (UTC)

artificial dna

Please, could this be integrated? (I am no expert): http://science.slashdot.org/story/09/08/18/0043212/Scientists-Learn-To-Fabricate-DNA-Evidence --91.37.35.170 (talk) 14:55, 18 August 2009 (UTC)

How to make working model on dna fingerprinting —Preceding unsigned comment added by 59.95.97.121 (talk) 07:50, 10 September 2009 (UTC)

I am not an expert either, but am an interested layperson, and I will try to put together a section on this matter. Anyone here interested in helping? Pv86 (talk) 19:53, 15 January 2010 (UTC)