Last year, California police used a new DNA technique to catch a criminal, the so-called Golden State Killer. In response to many of the privacy concerns that were expressed afterward, the Department of Justice developed a set of rules for when and how investigators would use these new techniques. Those rules were published two months ago.
Unfortunately, I don't think they address the most important issue – and neither has any of the news coverage.
First, a couple of notes and some background. Each person's genes can be thought of as the instruction set for building just about every part of their body. Most of the genes in each human are basically identical. That is why each person's body looks fairly similar. Some genes, however, have variations, which lead to differences such as eye color and skin color. Obviously, other things also affect something such as skin color – for example, sun exposure.
During typical human reproduction, a mixture of each parent's genes are combined to start a child. Some genes may change into a new variation, a mutation, during human reproduction, but the vast majority can be directly tied to one of the two parents.
Because of this very strong genetic connection between parents and children, extended family are also, somewhat more loosely, connected through their genes. Many companies have started in the last few years to analyze genes that are frequently different among people. Based on how many differences exist and where within a person's DNA, the company can then identify distant relatives.
In a long-standing, unsolved criminal case, California police and the FBI used DNA evidence they were confident came from the perpetrator. Police submitted the perpetrator's DNA to GEDmatch, a DNA-matching company, and identified several people who were equivalent to the suspect's third cousins. That connection was sufficient for police, through other work, to identify and prosecute a suspect.
My first reaction to hearing this story was to be impressed. I was impressed that someone thought of this creative way to use DNA evidence and actually got it to work. Solving crimes is important and, I presume, can be quite difficult. They should be congratulated.
As this story made the general news, however, it was clear that the more common first reaction was to be worried about privacy.
The recent Department of Justice guidelines are a response to this worry.
The guidelines limit such DNA searches to violent crimes and to identifying human remains.
Weirdly, police are required first to exhaust more traditional methods before using these DNA techniques. Police are also forbidden from secretly obtaining a DNA sample, say from a used tissue. They must have the person's permission or a search warrant.
Most of the news seems to have concentrated on whether this technique represents an invasion of privacy. Companies such as GEDmatch seem to have loose protections or requirements to keep DNA data from being used by police.
I think those concerns miss the major, new issue here.
As the first case in California demonstrates, using these new DNA techniques does not require a suspect's DNA. It is a fundamental fact of nature that I will always share a significant fraction of my DNA with my extended family. In fact, the power of this technique is exactly that we don't need a sample from every person. Just a distant cousin is sufficient.
At some level, my DNA is my business. However, do I have any privacy right to my sister's DNA? What about my cousin's? I honestly don't know the answer, but I don't hear anyone seriously discussing it publicly. The fact is that this technique will work extraordinarily well without a person's direct permission. Police only need the permission of your third cousin.
I could easily imagine a new DNA company advertising that it will help police solve crimes and requesting DNA submissions. I suspect most of us have at least one third cousin who would support such a project.
There have been rigorous studies estimating how many samples a company would need to be able to identify most people in the U.S. Assuming just a third-cousin connection is necessary, the answer is that samples from about 1.5 million people would cover about 60% of Americans of European ancestry.
With the ever-decreasing cost of genetic testing, these techniques are bound to become more attractive. In one sense, a way to solve crimes efficiently is a good thing. It is also really not clear how to think about privacy here. Just placing a wall between the police and these companies, or requiring explicit permission, does not address the main issue.
The main issue is whether or not I have a privacy right to my extended family's DNA. I don't hear anyone claiming that I can prevent police and a third cousin from collaborating. However, I am not sure most people are comfortable with it, either. As a society, I am pretty sure we need to pick one side or the other.
Christer Watson, of Fort Wayne, is a professor of physics at Manchester University. Opinions expressed are his own. He wrote this column for The Journal Gazette, where his columns normally appear the first and third Tuesday of each month.