Hacker News Reader

Comment by gph | Hacker News Reader

It makes me wonder if in the future there might be false convictions based on DNA evidence, and not from chimeras or inaccurate testing. DNA found at a crime scene has basically come to mean that person is guilty in the public mind (thanks CSI). Even though it's unlikely, it's possible to get some DNA on someone you're near in public who then gets murdered. So you've got DNA on the person, your cell phone is in the area, and you don't have a particular alibi. In most cases that probably means that person is guilty, but I can imagine a number of cases where the person could just be in the wrong place at the wrong time.

From the article it sounds like from where the DNA was located and other corraborating evidence that this guy is guilty. But I'm glad the article ended by saying some of the geneticists weren't condemning the accused based on the DNA evidence alone. I hope that type of thinking expands into the wider public.

View on HN

I wonder about this all the time.

Sometimes I comb my hair, or bite my fingernails. What if part of it fell on the floor and some random person steps on it. Stuck on the bottom of their shoe. A bunch of coincidental stuff comes up and my DNA is tested against a fingernail/hair under victim's shoe... match. I'm doomed.

Indeed, I fear the mentality that DNA-match = Guilty.

How many people have been sent to jail/put to death based on false DNA results or even accurate but not via the crime they're accused of.

> Sometimes I comb my hair, or bite my fingernails. What if part of it fell on the floor and some random person steps on it. Stuck on the bottom of their shoe. A bunch of coincidental stuff comes up and my DNA is tested against a fingernail/hair under victim's shoe... match. I'm doomed.

This is a plot point in the film Gattaca. I'd recommend watching it if you haven't already:

http://en.wikipedia.org/wiki/Gattaca

>Sometimes I comb my hair, or bite my fingernails.

You don't even need to comb your hair to leave DNA all over the place. A normal person will lose 50-100 hairs every day.[0]

Also, everyone sheds dead skill cells, and certain dermatological conditions can exacerbate matters.

[0] http://www.aad.org/dermatology-a-to-z/diseases-and-treatment...

...everyone sheds dead skill cells...

True. If you're in an office you haven't visited before and would like to be disgusted, examine the bottoms of a few mice.

From having sat on a criminal jury in the U.S., the opposite seems to be true: jurors wonder why the police didn't DNA test a small stolen object to "prove" it was held by the accused. There was no reason to run DNA (the object was found in the accused's possession, etc.) and it's not even clear that there would be any DNA to run (fingerprints would have been more likely), but it seems like jurors have come to expect CSI level stuff in every case. Instead, most cases seem to still be done on old fashioned eyewitness testimony and circumstance (the stolen goods were in your apartment, you were also around the place where they were taken...).

If only we could replace eye witness testimony: it's notoriously faulty.

We can, with high quality security cameras littered around public areas.

Most police forensics is junk science: http://lst.law.asu.edu/FS09/pdfs/Koehler4_3.pdf (page 4).

http://sites.psu.edu/dhlaw/2012/07/07/have-dna-databases-pro...

The problem with DNA evidence is that prosecutors like to jump to the end, the idea that a DNA match means that someone with DNA identical to that found at a crime scene has been found. In reality there is a long string of assumptions that must hold true in order for that supposition to also be true. The chain is only as strong as its weakest link.

>It makes me wonder if in the future there might be false convictions based on DNA evidence, and not from chimeras or inaccurate testing. DNA found at a crime scene has basically come to mean that person is guilty in the public mind (thanks CSI). Even though it's unlikely, it's possible to get some DNA on someone you're near in public who then gets murdered

Unlikely? It's very likely. We interact with tons of people each day. Also consider cases were the DNA is from a person close to the victim, but who isn't the murderer. Perhaps a friend came by the house and then left, and another came in and did the murder. Or a person sleeps with someone (perhaps an one-night-stand) and then is murdered, say by a jealous partner. The DNA could be from an interaction previous to the murder.

And of course, if the police is "sure" they have the suspect (like they often are, especially is he's say black or poor or something like that), they would also consider it "OK" to plant some DNA obtained for him to be "discovered" in the murder scene. Police has been found (in court) pulling such shit with falsified evidence) for decades.

I'm more concerned the Italian police were allowed to collect the DNA over 100 000 people in order to find the potential criminal.

People don't seem to realize how valuable and many times private DNA information really is otherwise a large number of them would have resisted this dragnet.

Am I the only one thinking this kind of large scale collection of DNA information should no be allowed without some very powerful circumstantial evidence.

Not just you happen to live in the same city a person was murdered in.

There was a series of murders in the north of England maybe 25 years ago that went unsolved. When DNA testing became available, the police essentially browbeat every male within plausible distance to be tested. They did solve the case. I'd have thought such questions would have been raised then, and perhaps they were.

Sounds like this case:

http://en.wikipedia.org/wiki/Colin_Pitchfork

Particulary interesting to me is the fact that the initial mass screening did not find the perpetrator, because he had paid somebody to give a sample for him. But he was eventually caught because the stand-in couldn't (or didn't want to) keep a secret.

Wikipedia mentions no controversy, although of course that doesn't mean there wasn't one.

Maybe people know, but don't care?

I wouldn't mind submitting a DNA sample if it helps the police find a murderer. Maybe it was some distant cousin of mine, and my DNA will help them narrow their search.

One of my former professors has done a lot of work on the issue of how DNA evidence is presented to juries. Basically, when you quote the Random Match Probability, jurors (and apparently journalists like the author of the article) basically forget about everything else: http://bioforensics.com/conference/RMP/RMP%20-%20irrelevant.....

There's two different logical errors involved. First, jurors conflate proving the assertion "this is the defendant's blood" with "the defendant killed the victim." Second, the "one in a billion" theoretical random match odds turn into more like "one in several thousand" when accounting for laboratory error. In fact, laboratory error totally dominates the error rate of the overall process.

Doesn't have to be accidental. The criminal could intentionally leave someone else's DNA at the scene. Or less subtly, leave a DNA Bomb [1].

[1]http://almost-human.wikia.com/wiki/DNA_Bomb

Just throw in more accurate phone/wearable tracking and you're in bigger trouble.

http://www.extremetech.com/extreme/126843-think-gps-is-cool-...

Reply on news.ycombinator.com