When there’s a criminal investigation, modern science often allows the police to compare DNA collected from the scene with that of various suspects.
It’s my understanding that scientists are usually looking for skin cells. I know they can get samples from things like cigarette butts. I have a case where they got DNA off a bottle of orange juice. Semen and blood work too.
When there’s a match, I’ve seen it explained as some astronomical chance that it’s somebody else (i.e. “there’s a 1 in 5 quintillion chance that this DNA matches somebody other than the suspect”).
What specific DNA is being tested? What sort of comparisons are being made? How do they do it?
Not a forensic lab person but I’m not sure what you mean by this question? As you said, it could be DNA from skin, hair, semen, etc. Are you thinking that DNA is different, if it’s from different parts of the body?
No, I’m wondering what part of the human genome is so particularly unique to each person that they can be identified with such precision. It can apparently be sourced from different parts of the body, but what is it that the scientists are looking for?
Forensic technicians process DNA drawn from body fluids and tissues (including skin cells, hair follicles, saliva, mucus, semen, blood serum, et cetera), use the polymerase chain reaction (PCR) to amplify (duplicate) specific DNA sequences, and looking at those sequences to identify unique short tandem repeats (STR), which are identified alleles (known variations in nucleotide sequences among human populations). By looking at the combination of unique STR sequences a person can be identified as being a closely related to the source of the sample, and often (but not always) uniquely identified. Sequencing and analysis of mitochondrial DNA (mDNA), which is conveyed exclusively on the matrilineal line, can also be used to distinguish close matches among paternal lines. Y-chromosome analysis can be used for the converse, since it is always inherited from the male line.
There is a standardized system in US law enforcement called the Combined DNA Index System (CoDIS) which indexes loci of known STR sequences for easy searching and stored in a national database. In the case of highly degraded DNA, other variable number tandem repeats (VNTR) can be used to perform direct comparison between samples and can provide highly unique profiles but require a greater degree of interpretation for a match.
DNA matches, when performed correctly, are essentially the most definitive form of personal identification possible. However, they are not evidence of a crime itself but just the presence of genetic material found at the scene. DNA contamination is common, both incidental (DNA is present at the scene for an explicable reason) and accidental (poor collection techniques munge external DNA with the sample) which can bring into question the integrity of the evidence. In addition, a number of forensic labs have been caught out performing inadequate analysis or testing on samples that were provided without appropriate provenance, so like all forms of forensic evidence handling the integrity of the chain of evidence is crucial to establishing that the association of evidence with the crime in question is valid.
I looked into this at one point. What happens is that one can buy assay kits for commercial PCR amplification machines which include all the necessary reagents, including primer DNA sequences for all the aforementioned CODIS loci, possibly plus some others. So the forensic scientist does not have to figure out each time what loci and alleles to look for; it is for the most part determined in advance and the procedures automated.
Basically, the way it works is that there are things called “restriction enzymes”, that will cut a DNA strand in every location where it sees a certain sequence. Do this to a whole strand, and you end up with a bunch of shorter strands. The specific pattern of strand lengths then provides the “DNA fingerprint”. This is super easy to do: High school students do it in classroom labs.
These days, though, if you really want to be sure, it’s not all that hard (though above the high-school level) to do a complete DNA sequencing: That is, read off each and every single base pair in the entire DNA. This isn’t usually done in criminal contexts, because it isn’t necessary, though it could be, for a sufficiently high-profile case, or a defendant with deep pockets.
Another application of this is to validate cell lines used in research. A vast amount of research dollars has been spent on research that isn’t reproducible because the cell lines used were not what they were thought to be. The famous HeLa cells are a well-known source of contamination, for example.
And it’s like fingerprints - the analysis of the patterns has been studied to death, and the statistics show that the odds of two samples at random matching gives a probability that becomes atraonomical very quickly.
For example, if A can be 3 to 12, that’s 1:9 odds. If B independently can be 2-8 (1:7) then the odds of both matching are 1:63; and so on. People get a random half from each parent, so odds are only half the DNA will match between siblings (except identical twins). Also over time changes creep into DNA, especially the “junk” between active genes. (Which is used to estimate population splits, for example how long ago did North American DNA split from Siberia?) So the less related two people are, the less likely all their DNA matches closely.
I read a study once where they said some forensic units use a fancy filter vacuum to suck up debris in a crime scene (like a closed room) to look for DNA samples from skin cells shed by recent vistors. One anomaly was that some people are “spreaders”, for some reason shedding a lot of skin flakes that seem to mgrate everywhere. They have encountered samples from people who could never have been even close to the crime scene - hitching a ride on other people’s clothing etc.
So like a fingerprint, all a DNA sample says is “yes, it’s this person”. What it means depends on what was found where.
In a nutshell, sequencing a person’s unique microbiome for identification. The scientist mentions that a person sheds ~36 million bacteria cells/hour. So even if someone is present at a crime scene and doesn’t touch anything or leave any trace of their own DNA, they may be identifiable through the microbes they leave.
One caveat I’ve heard is that a person’s microbiome is dynamic, so comparing them at different time points may be legally problematic. Are they still unique in that they are still useable even after a long period of time between comparisons? I don’t know the stats on it
I’ve seen quite a few articles on this but the actual scientific basis for asserting that an individual’s microbiome is a unique and exclusionary identifying method is…incomplete at best. You do host bacteria that will evolve to be best suited to the environment of your sinus and gastrointestinal tract, but making some kind of objective criteria that would stand up to legal challenge (especially by a jury of laypeople) is more challenging, notwithstanding the even greater problems with evidence contamination than with DNA when a potential suspect is just incidentally shedding millions of bacteria just by walking down a hallway or standing in an elevator for a few minutes.
I agree. I thought I’d post it because it may be of interest as related to the topic.
It makes sense to me that a person’s microbiome would be unique, but how do you definitively prove this? And they are dynamic. So how to you show that a microbiome sample from a crime scene matches a reference sample from a suspect?
However sequencing instruments, chemistries, and bioinformatics tools are always improving.
Contamination is an issue. Would the tech doing the sequencing have to run his own fresh sample each run as a negative control? Do it in a clean room?
And even if it’s doable, I agree DNA from public spaces would be problematic. But like a fingerprint on a murder weapon, if you violently strangle someone even while wearing gloves, you’re probably shedding a lot of bacteria on that person.
And like genealogy companies like 23 & Me that do SNP-chips for ancestry, some companies offer to “sequence” your microbiome, usually by a stool sample (I once interviewed for a sequencing job with a company whose business model was then to sell nutritional supplements. They seemed shady and I wasn’t interested after the interview. I saw a few years later they had gotten into legal trouble.) It makes you wonder if there will be a microbiome CODIS-type database in the future.
it alsowould seem logical to me that any close encounters of the human kind - or even touch-interacting with pets and farm animals - would likely make some changes to one’s biome. As would touching door handles etc.
I imagine crime scene microbiome DNA would come from shed bacteria living on the skin. A simple shower might not be enough.
Back when I took undergraduate microbiology, one of the lab exercises was to scrub our hands with antiseptic soap multiple times. After each wash, we would swab our hands and then an agar plate which would be incubated. I think we may have done the washes ten times. Even though microbial growth decreased with each wash, there was still a lot of stuff growing on that tenth plate.
It would be interesting to get a deeper understanding, but I suspect that our actual microbiome is pretty resistant to change. I found this, which is pretty useful.
I suspect new bacteria and fungi from surface contamination will have an uphill battle to establish themselves. Most won’t get the slightest foothold simply because they have never evolved to survive in the skin. The biota that live in our skin have a hard life, living off the meagrest of crumbs from our metabolism. Those that can survive are sitting on top an environment that is already well populated, with all the various niches occupied. Being able to reproduce enough to establish themselves against the existing population wont be easy.
The skin microbiome isn’t a surface population. They live within the epidermis - the layer of dead skin cells that forms the tough outer layer, protective of the vulnerable, live, dermis. Showering and the like makes essentially no difference - although obsessive cleaning can deplete the skin of oils, and that will affect the microbiome that depends on that energy source.
This is a symbiotic relationship. Our skin’s health depends upon a happy biome. It isn’t a passive population, but it itself actively managing the equilibrium. There are bacteria that actively kill unwanted interlopers. Also, the microbiome is different in different areas of skin. Exposed dry skin supports a different mix to moist areas, and oily areas are different again.
So obsessive cleaning can and does cause problems by messing with the microbiome equilibrium.
The paper above suggests that even after two years the microbiome population is pretty much unaltered.
What would be interesting would be to see how much the microbiome of couples overlaps. I’ll bet there is some overlap, but nothing like complete.
Am I supposed to be impressed? I had an … incident two weeks ago where I lost several TRILLIONS of bacteria in less than an hour. (Is there a Grotesque Medical Incidents thread for those who need to know more?)
