The Golden State Killer and DNA

Joseph DeAngelo, 2018 mugshot, alleged Golden State Killer

Unless you’ve been living under a rock for the past few days, you already know that the Golden State Killer has, it appears, been apprehended by:

  1. Sequencing DNA from the original crime scene
  2. Uploading those results to a genealogy data base to utilize techniques currently used for unknown parent searches to suggest or identify the killer
  3. Then, to confirm that they had identified the right person, discarded DNA from the suspect was sequenced which apparently matched the original DNA from the crime scene

I say “it appears” because remember, until he’s convicted, Joseph DeAngelo is still a suspect.

I have received more messages, texts and e-mails about this one topic than any other, ever. My phone has been buzzing like an angry bee with too much caffeine for days.

Unfortunately, in many news articles, the topic suffers from dramatic over-simplification at best and significant errors at worst. This combined with lots of fear stirs a toxic brew.

In almost all cases, the author writing the article clearly didn’t understand the subject matter at hand. Many leaders in the genetic genealogy community have been asked for comment. Having had more than one situation in which I was misquoted or my quote was taken out of context, I am discussing the issue in this article, where my comments aren’t boiled down to a one sentence sound bite. I don’t want anyone making a knee-jerk reaction with partial information. This topic deserves, and must receive much more discussion in a calm, informed manner.

There is a great deal of concern, curiosity, misinformation and incorrect assumptions in the genetic genealogy community as well as the media, along with emotions running at high tide.

I think it’s important to do three things:

  1. Discuss what actually happened.
  2. Discuss how genealogy versus both unknown parent and forensic searching differs from genealogy searching.
  3. Discuss associated concerns.

The Case

The Golden State Killer has been accused of at least 12 murders, more than 50 rapes and many burglaries primarily from June 1975 through May 1986. DNA evidence was collected, but DNA testing at that time had not progressed to the point where the culprit was able to be identified by utilizing his DNA.

A lot has changed, both in terms of DNA technology and other resources available since that time.

Last week, on April 25th, Joseph DeAngelo, now in his 70s, was arrested after DNA matching implicated him as the Golden State Killer. The news is ripe with stories, but this NPR article is a good summary as are the references at that bottom of the wiki article linked above.

Initial Concerns

Initially, two questions were being asked.

  • Which genetic genealogy company “cooperated” with law enforcement?
  • Did law enforcement have a search warrant?

As it turns out, the answer is that no testing companies “cooperated” and that no  search warrant was needed.

The next question was, “How safe is my DNA?”

Let’s talk about what happened, how it was done and how it affects each of us.


I was not involved with this or any similar case in any capacity, although I have been working the past few days to ferret out what actually happened, including discussing this privately and in public forums.

However, I am familiar with the techniques used as a result of my involvement with archaeological digs and ancient DNA, and I’d like to discuss what actually happened, as best we can unravel to date.

DNA Collection

At the time of the rapes and murders committed by the Golden State Killer, one police officer froze extra samples of the evidence, just in case, for the future. That future has arrived.

In the past few years, whole genome sequencing of ancient DNA and degraded samples has become possible. Probably the most notable are the Neanderthal and Denisovan genome reconstructions, beginning in 2010, but sequencing of forensic samples has become commonplace in the past few years.

From those ancient sequences, as long ago as September 2014, whole genome sequences were being reduced to just the DNA locations supported by GedMatch and the resulting compatible files uploaded there for comparison to other testers. This was possible because the raw data files are made available to testers by testing companies, so testers can modify the files in any way they see fit without the cooperation or involvement of any lab or company.

More ancient samples were added to GedMatch in the following months, and the ancient DNA comparison feature continues to be quite popular. No one ever thought much about it, but there is absolutely no reason that same technique couldn’t be used for other samples, and indeed, now it has.

Just 13 days before the arrest of DeAngelo, another homicide was solved by DNA sequencing. A murder victim, known as Buckskin Girl, found in 1981 was identified as Marcia Lenore King.

According to the non-profit Doe project, whole genome sequencing was performed, the file reduced to a format needed for GedMatch, and the file uploaded.

Again, there was no public outcry – possibly because a victim had been identified and not a criminal suspect, and because the event was not as widely publicized. However, it’s also possible that if the Buckskin Girl’s murderer left DNA evidence on the body, that sequencing could have identified both the victim and the murderer.

The identification of Buckskin Girl, however, did spur non-public debate within the leadership of the genetic genealogy field. Little did we know that the next case would follow dramatically in just two weeks.

GedMatch Matching

GedMatch is an open data base created in 2011 by two individuals in order to facilitate open sharing of autosomal matching between people, even if they tested at different companies.

Of the DNA testing companies, at that time, only 23andMe and Family Tree DNA provided centiMorgan information, recently joined by MyHeritage. Ancestry does not provide this information to their clients, so if an Ancestry client wants to see how they match other individuals in terms of actual chromosome locations and centiMorgans, they must transfer to either Family Tree DNA, GedMatch or now, MyHeritage.

Because GedMatch, with few exceptions during periods of change, matches customers from every vendor against customers from every other vendor, at least partially, they have become the clearing house for many people, especially Ancestry customers who don’t have the chromosome comparison option natively at Ancestry.

I want to be VERY clear about what you can and cannot see and do at GedMatch.

You can see your matches by the name they have entered, which can be an alias, along with their e-mail and how you match them. You CANNOT see the information of anyone you don’t match, unless you utilize another person’s kit number to see who they match. This has always been how GedMatch functions.

GedMatch users do NOT have access to your actual DNA file – ever. They can see who they match, and if they have your kit number, they can see who you match as well. Here’s an example of my own match screen.

Note – typically when showing GedMatch screen shots, I would blur the kit numbers and names in keeping with good privacy practices. However, since the point is to show you what one can actually see, I haven’t, because the top two matches are my own kits from Ancestry and 23andMe, and the third kit is that of my deceased mother whose kit I now manage. I also want to demonstrate that truly, there is nothing frightening or threatening about the information your matches see about you.

Best Matches

From a genealogist’s perspective, your “best matches” are to known close relatives, because when you match that relative and another person, especially on the same DNA segment, it’s a good indication that you share a common ancestor further back in time.

Genealogists build “clusters” of those types of matches in order to prove a relationship to a common ancestor. This is the heart and soul of DNA matching for genealogy.

For example, someone who matches you and your first cousin, both, on the same rather large segment assuredly shares a common ancestor with you and your cousin someplace in the past. The genealogical goal, of course, is to identify that long-deceased ancestor.

For example, if you match a first cousin, you know that your most recent common ancestor is one of your two sets of grandparents. Most genealogy matches are further back in time than either first or second cousins, making the identification of the common ancestor more challenging. Discovering that common ancestor is the goal of the game – because these matches to people with the same ancestor in their tree (generally) confirm that your ancestor is accurately identified. Some matches solve long-time family mysteries and break down brick walls.

However, not all brick walls are in the past.

Adoptee and Parental Search Matching

A few years ago, genealogists attempting to find unknown parents for adoptees and people with unknown fathers noticed that there were matching patterns to be followed successfully.

With millions of people having tested today, it’s much easier than it was a few years ago to find that key match (or matches) that reveals or confirms the identity of either an ancestor or an unknown parent.

While both genealogists and unknown parent searches look for close matches, the techniques diverge at that point.

Genealogists use a first or second cousin match to move backwards in time, looking for common distant ancestors.

In unknown parent searches, the same genealogical technique is used, EXCEPT, the person doing the searching could care less about older ancestors, such as great-grandparents. They are looking for their immediate ancestors – their parents.

Therefore, when an adoptee finds that critical first cousin match, they aren’t interested in figuring out a common ancestor for genealogy, meaning going backward in time. They covet that first cousin match for the purpose of coming forward in time, meaning towards the present in order to identify parents.

If you match to someone as a first cousin, you share a common set of grandparents. You can’t tell, without additional information, which set of grandparents, but given that you do match as a first cousin, there are only two positions the match can have in your family – either the pink or blue person above. This means that either your father or mother was a sibling to your first cousin’s parent.

You either share your father’s parents with your first cousin, or your mother’s parents, but you don’t know which – at least not yet.

With that much information, it’s fairly easy to uncover the rest. After all, you only have two sets of grandparents and anyone who is your first cousin will point to one of those two sets of grandparents.

You need to figure out who else matches you AND your first cousin, and then look at the genealogy of everyone who matches in that group until you discover the name of common family members/ancestors that you recognize, meaning an ancestor on either your maternal or paternal side to confirm that your first cousin matches you on that line.

Of course, for people who know their parents, figuring out first cousins is easy and takes about 2 seconds – but not so much for adoptees. Adoptees look to see how people who match them also match each other. For example, does the same couple or ancestor appear in the trees of multiple matches? In the example below, if the tester matches all three blue people as first cousins, the name of the blue cousins’ grandparents would be the same, suggesting that the tester’s grandparents were that same couple.

Next, it’s necessary to figure out which people who descend from the common set of grandparents might be candidates to be the parent the tester is seeking. In the example below, we’ve expanded the side of the three blue first cousin matches, adding their parents’ siblings as parent candidates for our tester. Factors such as age and location at the time of conception are taken into consideration when focusing on parent candidates.

If the tester doesn’t know who their parents are, they would be VERY interested in determining ALL of the children of the grandparents of their first cousin. Because one of the children of their first cousin’s grandparents IS THEIR PARENT.

In our example above, let’s just look at one of the grandparent pairs of the blue first cousins. The first cousins know who their grandparents are. The tester does not. In this case either the father or mother of the tester is the child of the first cousin’s grandfather and grandmother. Meaning that the red mother is the female child of the grandparents, or the green father is the male child of the grandparents.

We know that the grey parents of the first cousin matches can be eliminated as the tester’s parent. If the first cousin’s parent was also the parent of the tester, then the first cousin wouldn’t be a first cousin, but would be a full or half sibling.

However, the matching first cousins’ parents have three siblings who have not DNA tested, nor have their children, shown in pink and green. One of those three siblings IS either the father or mother of the tester. Of course, if the grandparents didn’t have any female children, then the tester’s father is one of the green male children of the grandparents, and vice versa.

In the example shown below, the tester’s mother IS the female child of the grandfather/grandmother pair and has been moved into place. This would be determined either by direct testing of the pink or green people, or their descendants, or by process of elimination through DNA tests of the other siblings or utilizing other pieces of information such as age and proximity.

Some adoptees are lucky enough to test and discover that a parent has tested and is waiting for them. Sometimes an unsuspected half sibling appears. Sometimes, there is no close match and the adoptee has to do more research work, including tracking people through social media and other means to find candidate family members to DNA test or to see if they know who might have been the much-sought-after parent.

Search Techniques

This type of research work has been taking place for years, individually, through groups like DNAadoption and DNADetectives who utilize volunteer search angels, as well as by several researchers who make a living doing this type of personal search. My focus is not on adoption search cases.

No one has seemed to consider this unethical, even though some of this work, especially when a parent isn’t immediately evident, involves utilizing the DNA of the tester’s matches and their matches’ relatives, connecting the family dots through social media, specifically Facebook pages, to discover the identify of someone who may not welcome that discovery. However, like GedMatch, Facebook, while not intended for this purpose is public and is heavily utilized by adoption searchers.

Some adoption search cases end very well – with heartfelt beautiful reunions welcomed by all parties. Others not so much, potentially upending the life of the biological parent that was established after the adoption took place which leads to a rejection that devastates the adoptee. Much of the damage can be done by the search process itself, meaning that the biological parent is “outed” by the process of people working through relatives who have tested and match in various ways. Of course, they ask questions to identify the biological parent – meaning that by the time the parent is identified they have no say about their own privacy.

Once DNA is uploaded to a data base, the search techniques for biological parent searches and to identify Buckskin Girl and DeAngelo, are exactly the same.

These searches all utilize matches to others, and the matches’ trees, to move forward in time to current to search for contemporary people, not ancestors further back in time.

Back to the Golden State Killer

Ok, back to the Golden State Killer.

We have the killer’s DNA sequence from the original crime scene and the file reduced to the number of DNA locations utilized by GedMatch.

Someone, presumably one of the investigators working on the case, uploaded that file to GedMatch, which appears to be entirely permissible because the police have legal custody of that DNA sample.

Let’s say the investigator, just like a genealogist, found a first cousin match, or even more distant (read difficult) matches further back – and they did exactly what people searching for unknown parents do. The investigator eventually worked through all of the possibilities based on common matches – then looked at age, location, opportunity and factors that might exclude some candidates. In this case, because it’s a rape case with the criminal obviously a male, females would be excluded, for example.

Evidence from DNA matches to the biological sample of the Golden State Killer caused the police to focus on DeAngelo.

After DeAngelo was identified through matches as a suspect, the police obtained his discarded DNA. Discarded DNA could be anything from a coffee cup thrown away to a cigarette butt or something from the trash.

That discarded DNA was sequenced, and a few days before his arrest, uploaded to GedMatch as well. The discarded DNA apparently matched the earlier sample from the killer as “himself” and the other people that the killer matched in the same way – establishing the fact that the Golden State Killer and DeAngelo were one and the same person.

You can see that I match my own 23andme and Ancestry kits as my closest matches in the GedMatch example I showed.

In essence, what the DNA of “the killer” obtained from the crime scene did was to generate leads through matching that allowed the police to identify DeAngelo and obtain a sample of his discarded DNA in order to verify that DeAngelo was the same person as the killer. Of course, he’s still a suspect today, not yet convicted.

Cooperation or Search Warrant

The police, in this case, didn’t need to ask for anyone’s cooperation. They already had the sample from the killer, they did what hundreds of thousands of others have done and simply uploaded the file to GedMatch.

The investigators didn’t need a search warrant because they weren’t asking for anything from GedMatch not already freely given, meaning matches to anyone who has already uploaded their information.

The investigators only used that matching information to generate tips for further investigation. They repeated the entire process with the discarded DNA sample to verify the earlier results obtained with DNA from the crime scene.

It bears noting here that if DeAngelo’s DNA had NOT matched that of the killer and the other people in the same way the killer’s DNA had matched them, then the discarded DNA would have eliminated DeAngelo as a suspect.

So, no genealogy testing company had to cooperate with anyone, nor was a search warrant necessary.

What’s the Rub?

We now have a monster about to be brought to justice. Two weeks earlier, Buckskin Girl, a murder victim, was identified and the family will finally have closure, 37 years later. Both of these are unquestionably wonderful outcomes.

So why are some people upset?

In some cases, people are simply confused about the process involved, and they will be relieved when they understand what actually happened – that their DNA was not “handed over” to anyone.

Some people have broader reaching concerns about privacy.

It appears that the word “police” combined with the word “criminal” caused a great deal of fear and trepidation, especially since a suspect was identified this time, not a victim and not someone’s biological parents.

Some people don’t want their DNA utilized to identify a family member, no matter what that person has done. And yes, that’s very nearly an exact quote from an e-mail I received.

Others are simply uncomfortable with their DNA being used in any kind of a potential criminal setting – even to identify a victim like Buckskin Girl.

One person says that it just makes her feel “creepy.” Oddly enough, that’s how I feel about Facebook now.

If you think it’s fine for adoptees to identify parents using these techniques, but you don’t think it’s alright for victims or criminals to be identified, I’d like to ask you to consider the following scenario.

A underage female is raped and becomes pregnant. She reports the rape to police at the time. She opts to have the child instead of having an abortion, and the child is placed for adoption. The rapist is never caught, and the young woman goes on to establish a new life and marry, not telling her husband or children born to the marriage about the rape, or the child placed for adoption. The expectation of the mother at that time was certainly that “no one would ever know,” whether those words were ever in an adoption contract or not. The fact that adoptions were (and still remain in many places) closed speaks to the expectations set for the mother.

Years pass, and today the adopted child, now an adult, tests. Both of the adoptee’s biological parents are identified through matches to relatives of the adoptee’s parents who have tested, such as first cousins in our earlier example. The adoptees parents themselves did not test.

Results were:

  • The life of the mother, a victim who did nothing wrong or illegal, and who chose to give the child life, is upended through the process of being identified.
  • The father who is a rapist, a criminal, is also identified.
  • The adoptee is subsequently very unhappy with both results for different reasons, but cannot press “undo.”

I’m NOT inferring that these data bases shouldn’t be used for identifying parents. I AM saying that we need to consider that the techniques for identifying parents, victims and criminals are the same. The outcomes are not always positive in parent searches AND these areas are or can be incredibly intertwined. Unraveling or prohibiting one effectively prohibits others. How do we treat everyone fairly and how are those rules, whatever they might be, enforced, and by whom?

In other words, how do we “do no harm”? After all, this started out to be genealogy, a fun hobby, and has now progressed gradually through a slow crawl to something else. Here we sit today.


In the example rape case above, neither the biological mother nor the father had tested, but their family members had – just like in the Buckskin Girl and the Golden State Killer cases.

Today, relative to the Golden State Killer, people are upset because the database, GedMatch, into which they uploaded their DNA file for genealogy was used for other purposes – specifically to apprehend the Golden State Killer. They feel that isn’t the purpose for which they uploaded their DNA.

Any one of us could have been one of the matches to the Golden State Killer and some people obviously were. It bears repeating here that no one’s DNA or results were “handed over,” and the only people affected in any way was someone that matched DeAngelo, and probably then only the closest matches. Many time people’s trees are utilized and their cousins never contact them, so it’s certainly possible that people who match DeAngelo have no idea still to this day.

The usage evolution for GedMatch from genealogy to other functions has been a slippery slope, although clearly no one realized at the time, when several years ago uploads began with modified ancient sample kits. Later, people began to use the GedMatch database (among others) to identify biological parents, then victims and now criminals.

Other people feel that searching for parents is genealogy, but identifying criminals is not – even though the search techniques are exactly the same. In our rape example, the mother who was a victim was identified and the criminal rapist father was identified as well by the same DNA test. The tester’s intent was only to reveal their biological parents – hoping for a loving, tear-filled reunion. That’s not what happened. The process of finding their parents also revealed the associated circumstances.

You can’t separate these usages into separate “boxes” anymore, because they overlap in unexpected says. That rape case wasn’t hypothetical.

I have absolutely no sympathy for the rapist, in fact, quite the opposite – but I feel incredibly bad for the young mother who has now been twice victimized. First by the rapist and second by the process used to track her, through relatives who began asking lots of difficult questions.

Last fall, in a Facebook group I follow, I was utterly horrified to see someone post that in the adoption cases she works, she encourages the adoptee, when they feel they are “close” to identifying a parent, to send registered letters to all of the family members, asking them to test, hoping that those who aren’t the parent quickly test to absolve themselves and as a way to flush the parent out.

It’s Not Just Your DNA

In either case, the DNA of the RELATIVES of the person being sought, be it a parent, a victim or a criminal, is what is used to find or identify the desired person. People who have uploaded to GedMatch are now concerned that they might be that relative whose DNA is used in a way they did not originally anticipate. They are right, and not just about this particular criminal case – but about the many types of usages other than strictly genealogical that looks backwards in time.

Perhaps the people who uploaded never thought about the fact that their DNA is/was being used for adoption or missing parent searches – or perhaps they are supportive of that activity. Maybe they thought that identifying victims, such as Buckskin Girl was a great use of the data base by investigators. Maybe they never thought about the fact that searching for criminals who leave DNA specimens behind uses exactly the same research and matching techniques as adoptees’ parent searched.  Perhaps no one stopped to think  that the same search can identify both parents, a victim and a criminal at the same time.

Maybe they were naïve and never thought about it at all or didn’t read the GedMatch statement that said (and says):

In today’s world, there are real dangers of identity theft, credit fraud, etc. We try to strike a balance between these conflicting realities and the need to share information with other users. In the end, if you require absolute privacy and security, we must ask that you do not upload your data to GEDmatch. If you already have it here, please delete it.

I can’t tell you how many of the posts and e-mails I’ve seen about this topic include the word “assume,” and we all know about assume, right?

Maybe, like me, some people have thought about that potential situation and want criminals, regardless of whether they are relatives or not off the streets. If they are relatives, so much the better, keeping my own family safer.

Some people may have been uploading their relatives’ DNA samples to GedMatch or any other site other than where the relative originally tested without the relative’s permission. If that’s the case, the person either needs to obtain permission, pronto, or delete the person’s DNA they uploaded without permission.

GedMatch’s Statement

GedMatch has posted the following statement.

Testing in the Future

Another concern voiced this week is that people, especially relatives that we want to test, will be much more reticent to test in the future if they think the police can “take” or “access” their DNA. That’s probably true, so we need to be prepared to explain what actually happened, and how, to eliminate misconceptions

However, it is true that DNA in these databases has been and is being used for things other than genealogy. This is also the purpose of informed consent – with an emphasis on informed. Bottom line – the cat’s out of the bag now. Perhaps these incidents together, meaning parent searches, the identification of Buckskin Girl and the arrest of the Golden State Killer, will bring home the warning that was previously noted on GedMatch.

If you’re not comfortable – don’t upload. This also means that people MUST STOP simply telling other people to upload to GedMatch as a cure-all for everything that ails genealogists. If you are making the recommendation, you also bear the responsibility for full disclosure or at least a caveat statement.

“GedMatch is great for genealogy matching to each other across vendor platforms <or words of your choosing>. It’s also used for adoptees searching for their parents, was used to identify Buckskin Girl and played an important role in the apprehension of the Golden State Killer.”

As a result, GedMatch now provides a way to remove your entire account, if you so wish. GedMatch needed to do that for GDPR anyway. As long as we are on the topic, GDPR, which goes into effect on May 25th tightens privacy significantly for any vendor or company that includes records of any UK/EU resident. You can read about that in my articles here and here.

Every (major) testing company, along with GedMatch provides the option of removing your DNA results if you are so inclined.

As for people being hesitant to test, certainly some already were and some will be. But there will also be others that only first heard about genetic genealogy this past week and this notoriety won’t deter them one bit. Some people will actively choose to participate, knowing that they can later change their mind if they so choose. I notice the GedMatch site has been busier than ever.

In summary, the police did not “take” or even ask for anyone’s DNA. They simply uploaded the DNA results of a criminal, taken from the crime scene, and looked at the matches generated in order to make an ID, at which time they obtained the DNA of the suspect which matched the DNA from the crime scene.

Just like genetic genealogy, DNA without supporting evidence won’t be much good, but now they have someone identified to work with, collecting other evidence. Where was he? Does the DNA at multiple scenes match his? I would think in terms of a prosecution that these matches and arrest is only the beginning, not the end of the process.

Given that none of the major genealogy companies cooperate with law enforcement without a search warrant, it’s a WHOLE LOT easier to obtain your discarded DNA than to obtain a search warrant. Furthermore, there is no chain of custody with DNA from a genealogy data base, but there certainly is from a rape and from a discarded cup. If the DNA of the criminal from the scene, and the suspect’s DNA from a discarded item match as the same person, that’s pretty conclusive and damning evidence.

Of course, fear begets fear and the old questions of government access and other issues bubble up again.

Another question I’ve received is about whether the usage of GedMatch for the Golden State Killer case opens the door for DNA to be obtained by insurance companies. First, you’d have to test and upload something. There is nothing to “get” if you don’t – and the insurance company would need a search warrant (and probable cause of a crime) to retrieve your DNA from any testing company.

GINA legislation protects American’s today from discrimination when obtaining health insurance, but it doesn’t extend to life and other types of insurance. However, when I applied for life insurance some years ago, they took a blood sample and if I wanted life insurance, I had to authorize whatever it was they wanted to test in that sample. I’d wager that today, they would run a DNA test in addition to checking for other health indicators. No GedMatch or testing company is needed or desired – in fact – an insurance company requires chain of custody which is why they send someone to your house to draw your blood.

What To Do?

What you do with your DNA sample is entirely up to you. Everyone will make their own decision based on their own circumstances and preferences.

Some people have removed their DNA from the various databases and in essence, have stopped participating in genetic genealogy.

Some have made their kits at GedMatch either research or private. Research means that you can run the kit and see matches, but others can’t see you. That certainly defeats the spirit of collaborative genealogy.

Some people have evaluated the evidence at hand and have made the decision to continue as normal – just more aware of other uses that can, have and may occur.

This story and others similar will continue to arise and unravel, and many questions will likely be asked and hotly debated over the next many months and years, both within and outside of this community. I would not be surprised to see legislation of some type follow – which has been one of the biggest fears within the genetic genealogy community for years. Legislation by people unfamiliar with the topic at hand will likely be overreaching and extremely restrictive. Let’s hope I’m wrong.

Like many others, I’m concerned that the genetic genealogy field will become a victim of its own success. I hope that doesn’t happen, but at this point, the cow has left the barn and that door really can’t be effectively shut. All we can do is to be transparent, make informed choices, assure that we have the consent of anyone whose kit we manage and to advocate for sanity.

My Decision

I’ve made my personal decision and my thought process worked like this:

  1. I haven’t done anything that I need to worry about.
  2. If a family member does something they need to be arrested for, I hope my DNA helps.
  3. If I were the family of the victims, I would want them identified AND their murderer/rapist put away forever. (Disclosure, I have had a family member raped and a different family member murdered.)
  4. As a citizen, I want criminals such as rapists and murderers identified and removed from society through any legal means possible.
  5. DNA testing also exonerates people who were wrongfully convicted through advocacy groups like the Innocence Project.
  6. DNA eliminates potential criminal candidates as well as pointing the finger directly at others.
  7. Using the techniques utilized for unknown parent searches, an identification is seldom made as a result of ONE match only, unless it’s immediate family. Therefore, if you remove your own DNA from the data base(s) for matching, your cousin and their cousins are still there – so your criminal family member’s goose is still cooked. It might just take a little longer in the stew pot.

My DNA stays online and I continue to support all of the major DNA testing companies that provide matching and accept transfers, including GedMatch.



I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.

Thank you so much.

DNA Purchases and Free Transfers

Genealogy Services

Genealogy Research

144 thoughts on “The Golden State Killer and DNA

  1. Your comparison with Facebook is a good one.
    People don’t seem to be afraid to use it to discuss this issue, despite the recent adverse news.
    And that access was in hundreds of thousands of cases at least, not one.
    And presumably it is being used in heaps of legal cases – without court orders and without having to find some scrap of evidence at a crime scene to enable a match – fingerprint, DNA or whatever.

  2. Thank-you for a well written article, for me you took a complex subject and presented it in an easy to understand format. I whole heartily support your seven points at the end of this article.

    On a related subject “DNA Ownership” would it be safe to say that I own the DNA of a deceased relative if it was on a letter written to me by them?

  3. Actually, Ancestry DOES have a feature where you can view how many centimorgans you share with a DNA relative. It’s not as easy to locate as on other websites, and there is no ability to sort your DNA relatives into different orders based on DNA, but you can view base centimorgan numbers.

    Just click on the profile of a DNA relative and click on the little grey “i” button underneath the predicted relationship section. That will show you how much shared DNA you have and how many DNA segments you match on.

  4. Roberta, excellent blog. One small correction – “Just 13 days before the arrest of DeAngelo, another homicide was solved by DNA sequencing.” We at DNA Doe Project solved the identity of the victim, but not the homicide. (If only!) That investigation is ongoing. But otherwise, a superb job. Just the sort of clarity, articulation, and absence of fear-mongering that is needed as we all work to adapt to this new world. Thank you! – Margaret Press, DNA Doe Project

  5. Thank you for this wonderful article. I know some who have all of a sudden turned skittish about testing. We have now identified my husband’s biological parents, and I am so thankful for that ability. They are both deceased, but the family members have welcomed us with open arms. And I would gladly volunteer my DNA to help identify a criminal.

  6. Thank you for such a well written piece on this subject. I hope many will read it and benefit from it in order to form their own decisions instead of relying on FB comments and group reactions to draw their conclusions.

  7. Thanks for your explanation.
    Here my consideration
    1. I don’t remove my dna from GedMatch
    2. I think living people have to choose by himself to have dna test (so i could not request for my Young sons)
    3. I’m afraid in case of potential misuse of discarded dna to build a false crime scene. I think dna it’s a strong ‘tool’.

  8. Thank you, Roberta. You’ve provided a much needed “cooling breeze” of rationality and commonsense. I’ve made sure to bookmark your article for future reference!

  9. There comes a time in everyone’s life when we’re faced with that question that generations before us have had to answer. Do I want to be part of the solution, or do I want to be part of the problem? Criminals beware, my DNA results will remain uploaded at GEDmatch.

  10. Cousin Roberta, Thank you so very much for this article. I spent almost all weekend trying to get folks to understand the points you so eloquently stated. BRAVO! and THANK YOU!
    In regards to adoptees discovering they were the result of a rape has always bothered me. But, then the need of those searching is so overwhelming that soon passed. Also, I think, the rate of births due to incest and rape are minuscule compared to births by consenting couples. Not to say those are not important.
    I would also like to say that sometimes the truth coming out is part of the healing process as painful as it might be. I know that is true in the case of many adoptee birth parent searches and results. There are many reasons birth mothers, fathers and adult adoptees suffer.
    I also think the laws for access to original birth records needs to be changed in 41 states. Perhaps this is the seed that will do that. I doubt it, but maybe. Genetic Searching does not give those involved much Privacy (that is a whole other discussion). I might add, the Constitution of the United States does not provide for any guarantee of Individual Privacy.
    One of the facts we try to explain to legislators, in changing the law for access to original birth records, is that using DNA, contacting 3rd/4th cousins (as in the GSK case) or closer, opens the door all the way up to the birth parent(s).
    I think a lot of this hoopla has been over misinformed reporter comments and frankly, even within our own community, those that just like to argue!
    Thank you,

    • “The Constitution of the United States does not provide for any guarantee of Individual Privacy.”

      What about the 4th Amendment to the Constitution? The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.

      • No it does not Guarantee any individual privacy rights. I believe this issue went to the U.S. Supreme court and Justice Sandra Day O’connor sent this case home after a year and a half. When the Plaintiffs, who brought the suit, could not come up with a good Constitutional reason. This particular case had to do with keeping birth records sealed after Measure 58 passed in Oregon. I believe it was the Doe v Sundquist case.

  11. I learned something today. I had always thought that the discarded DNA was put into CODIS because I assumed the original sample from one or more of the crime scenes. And it would be easier for the detectives as they were used to using CODIS. I never thought of the dna from the discarded sample being put into Gedmatch.

  12. You know there was a lot of work that was put into those family trees. It would be a shame if they were not made available to the families themselves for their own research.

    • About a decade ago, I attended a presentation where the guest speaker was employed by the Department of Defense to look for the closest living descendant of a deceased military member. This information was needed for two reasons, closure and disposition of the remains. They provided her a letter authorizing her complete access to all governmental records. However all her research belonged to the DOD and could not be shared with the living descendants. Her work around, if asked, was to inform the living descendants where she found the records.

  13. I have contacted several cousins in Europe about the GDPR. They are less concerned about than we are. So I wonder how it will affect genealogy. They have shared some info with me and looks like would not be a problem. I was concerned on how it might affect me trying to find living cousins. I will have to wait and see.

  14. I knew all along about adoptees finding out they were a product of rape. However I had never thought about when they look for parents, the one could be the rapist, the criminal. Well I did but not in the context you presented. And how that is really very similar to what is occurring now with criminal investigation. Everyone was ok when it was an adoptee doing this but for another circumstance they are not. When DNA testing started it really was more for genealogists. But then after time adoptees starting using. Lots of adoptees then tested and businesses was created. The adoptee searching has been a big influence and sometimes overshadowed those that are trying to do just old fashioned genealogy.

  15. Thank you. Your adoptee example is our story. Mom died in 1998 along with what happened in 1943. Was she the merry widow or a victim, we will not know. We have a sister and we think we have the fathers family.

    A fan of Gedmatch!

  16. Thank you for this clear explanation. I’ve been explaining “chain of custody” and “insurance companies who want your dna will get it directly” on Facebook, but I’ll share your post with my many cousins who have tested for the more complete description you’ve given about the process and results. Like yours, my many uploads will remain at Gedmatch. Before I upload more, I’ll refer cousins to this article. Two cousins who tested at Ancestry may choose not to participate further, that should be their choice.

  17. This has become a hot button topic.The expressed opinions of some appear to be anti-law enforcement, vocal and critical of the police no matter what they do, or whomever they bring to justice – even if the means that law enforcement used were legal.

    It’s easy to play “devil’s advocate” or aspire to act as a self-appointed public defense attorney, striving to find any possible fault with law enforcement. However, most people will agree that no matter how this suspect was initially found, he now needs to be brought to trial and held responsible for his alleged numerous and heinous criminal acts.

    To paraphrase Joe Stalin: A single death is a tragedy. Numerous deaths are…(fill in the blank).

  18. I’m personally very happy that DNA was used to find this scumbag. I don’t want to be disrespectful of other people’s concerns or opinions, but I really think people need to get their priorities straight. In the unlikely event that I have any relatives who are serial killers, I pray to God (or would if I were religious) that the police can use my DNA to catch them.

  19. Thanks for the detailed report, I read articles around “mainstream media”, but there were little details..

  20. In other words, the police did the same techniques as what the normal genealogist does and they came up with a serial killer – maybe just a bit more feet on the Street after they determined a family which the killer belonged to.

    For, they encourage you to share your whole 25,000 list of cousins with a newly found DNA match to figure out if there are any common matches – their other tools such as surname searches is a joke and incomplete stopping after say 60 pages or 3,000 names. There is no thought of them developing a technique to share 30 pertinent DNA Match names which can be checked another person’s results list.

    If you invite another person to view your results, you are opening up your life and cousins to a complete stranger.

    And mentioned adoptees is a red herring. All of us doing genealogy are doing the same type of searches for unknown parents OR gr-gr-gr-grandparents. It makes it sound that only adoptees should be afraid of the techniques rather than every one of us.

  21. Thank you for being the voice of reason and explaining things in a precise and clear way.

    Terry Moore _____ Terry Moore, CG 6001 Tarnhour Court Raleigh, NC 27612 919-649-5723

    Certified Genealogist and CG are service marks of the Board for Certification of Genealogists used to identify its program of genealogical competency standards and used here under license.

    On Mon, Apr 30, 2018 at 10:56 AM, DNAeXplained – Genetic Genealogy wrote:

    > Roberta Estes posted: ” Unless you’ve been living under a rock for the > past few days, you already know that the Golden State Killer has, it > appears, been apprehended by: Sequencing DNA from the original crime scene > Uploading those results to a genealogy data base to ” >

    • Terry, it is not appropriate for you to advertise your services on this blog…………Think about it.

  22. I’m all for DNA helping to shine some light on the past, for everyone searching for someone. I feel bad for victims who may have wanted to leave the past behind them, but offspring deserve the truth if they are willing to search for it.

  23. Thank you for a clear explanation of how genetic genealogy works. It may seem straightforward to those of us who have been using it for years, but it is complicated for the novice and most reporters facing deadlines seem to be struggling to understand it.

    I was an early adopter and asked my father to do a y-dna test as soon as a test became commercially available. When autosomal testing was offered my parents, a great uncle, and I were tested at FTDNA. I received their permission to upload our results to Y-Search and GEDmatch to increase our odds of matching. Today I deleted our results from both services.

    While I’d be thrilled to support LE in capturing violent criminals like DeAngelo, I’m concerned about a lack of oversight from the courts. What if LE decided to use my family’s DNA and my 30 years of research to identify individuals involved in lesser or nonviolent crimes like the one in five who now sit in prison for nonviolent drug offenses? Without warrants, without judicial oversight, I’m just not comfortable having my data so easily available to LE.

    Thanks for addressing this issue. It’s a fascinating topic and something I’m sure we’ll be debating for some time. It will be interesting to see how it plays out in the courts.

      • Richard, this document refers to the use of governmental databases, not open source databases like GEDmatch. LE in California had to receive approval from an interdisciplinary committee to search FDS which they did initially and had no hits. That’s when they turned to GEDmatch without a search warrant with no judicial oversight and without even contacting the owners of the database. I find that troubling.

      • While they may not have had judicial oversight, they had to have DNA and lots of investigator time, which lessor crimes just don’t get. Law Enforcement is harsh on drug offences because they are easy, often the drugs themselves are all the evidence necessary for a conviction. I’m all for reforming our drugs laws to make more sense, but I’ll believe LE using GED match for drug crimes when it happens, which I just can’t imagine with the current state of our society.

        Regardless as to how any of us perceive it, we all have the ability and right to remove our data at will.

      • I know that here in Oregon they do use DNA testing for some lesser crimes. The state lab recently had to put the processing of DNA related to property crimes on hold because in 2016 the legislature mandated that they catch up on over 5,000 unprocessed rape kits. Five indictments have come from the processing of these old kits. That they sat gathering dust while the lab was processing DNA from property crime raises some question about LE priorities at least here in Oregon and just adds to my sense of unease.

        I do agree though that, while DNA testing for all kinds of lesser crimes is possible, it seems unlikely that they would have the resources to to solve them using GEDmatch in its current form. DNA circles like Ancestry uses would make it somewhat easier.

    • Yes, LE uses codis (based on MT DNA) for lots of things, it’s most effective with repeat offenders. It seems that many LE agencies need more outside guidance on what their priorities should be. Unfortunately sometimes even the outside guidance is far from optimal. Some elected officials think they know best rather than listening to all of their constituents.

  24. Thanks for an excellent and most informative article! I’m curious to learn more about ID’ing first cousins. You mention discovering a first cousin via GEDMatch, and from there identifying the common grandparents. This raises two puzzles:
    1) What measures lead one to think, yup, that’s a first cousin specifically (as opposed to, say, an aunt, a cousin once removed, or an unknown sibling’s child)? Are there benchmarks for centimorgans or other measures?
    2) Certain relations would seem to be at a similar distance. For example, a cousin is 4 “steps” away, two up to grandparents and two down again. But an unknown sibling’s grandchild, or a great-grandparent’s child other than your grandparent would also be 4 “steps” away. So, is there a way to look at GEDMatch results to distinguish a first cousin from someone with a relationship at similar genetic “step” distance? Presumably inquiring about age would resolve this, but short of making contact, are there clues to be found using GEDMatch alone?

  25. Great article and well-explained! Thanks! You did an excellent job in explaining how the identification of DeAngelo didn’t violate any laws and was not out of the norm of what happens virtually every day. I must admit I didn’t read every word of the article or comments, so please forgive me if I’m making an observation that has already been asked and answered … what strikes me is that while you did a great job on the identification part of the DeAngelo case, you didn’t discuss what comes next, the process of CONVICTION. It seems to me that the DeAngelo prosecution team will have a substantial 4th Amendment hurdle to overcome in order to effect a conviction. All the defense will have to do is to convince one member of the jury [or any appellate court] that the GEDMatch search was an overly broad “fishing expedition” and DeAngelo walks. If the search is ruled a “fishing expedition” then any and all evidence found as a direct or indirect result of that search will be deemed “fruit of the poisoned tree” and will be ruled inadmissible. Included in that indirect poisoned evidence would be any discarded DNA left behind by DeAngelo. While I have not the least bit of empathy for this monster, I do have empathy for the rule of law. It will be very interesting how this trial proceeds.

    • Well said, Chris Rivers! This case will most likely raise questions about warrantless searches, and rules of admissibility for evidence. If I were the defense counsel, I would definitely raise those constitutional questions.

  26. I’m an attorney. I live in Alabama where things are certainly not perfect. I’m not at this moment practicing law, and I’m not at this moment giving any advice. I’m making an observation: A lot of people are talking on Facebook and in other forums like comments after “news” items like they are the all-knowing resource for all things legal, moral and proper. I’ve noticed that even lawyers are making value judgments!! My request is simply that we genetic genealogists (professional as well as hobbyists) just put all this to bed. I doubt anyone will sue you. Unless you’re a licensed attorney in California, probably you should quit thinking about what’s going to happen to this case everyone is talking about. Enjoy the fact that your DNA matters. It does. It’s stellar. We have the stars in our veins. On the other hand, this situation reminds me of how stupid people are, generally. Let’s just have fun with our genetic genealogy. Go for it!! And, let it go.

      • Certainly! I ended on kind of a silly note though…maybe take that part out about “we have the stars in our veins,” ….a little over the top lol! I was just ending on a thoughtful note that we all collectively have a stake in our collective DNA. Those little bits of amino acids are just arranged a bit differently in different people.

        • We do have the “stars” in our veins, ancient stars and future stars. They are small but brilliant strands of what we call DNA! They connect us to the past and will live in the future for ever and ever as long as there are descendants.

  27. Thank you for the article. Well written especially the last part as that was my decision as well.

    I’m looking for my father. I feel like it’s a needle in a haystack though. I’ve got no name. I do have knowledge about things such as eye color, body build, allergies, etc. My father committed heinous criminal acts. I saw him do them. I’ve reported the crimes to the cold case division in the state where they took place.

    I had planned to submit my DNA to GEDmatch until the arrest of the alleged Golden State Killer occurred. That freaked me out big time. In the end though, I decided to follow through and submit my DNA to the site.

    I hope by adding my DNA to there that he’s found and arrested.

  28. I’m hearing people say that it would have been difficult to get the suspect’s DNA into the databases at Ancestry and 23andMe. I’m not convinced that this is true. Wouldn’t it be relatively easy to prepare a DNA sample suitable for analysis at these companies?

    • We all shed DNA. But the offender does too. Plus, the location, context and quantity is relevant too. For example a rapist deposits DNA in semen in the vagina of the victim. Pretty difficult to set up fraudulently by the rapist.

  29. Thank you for the great article and explanation. If this is the way to take down creeps who commit these horrible crimes, then hallelujah, let’s do it.

  30. I would think that the victim would be pleased that the rapist was found out.
    As far as children with unknown parents, it is natural and perhaps good for their well being to find their parents. Its a who am I , really , who am I question. It would drive me crazy.
    It happened in our family, child out of wedlock. Father waited to hear if there were results. He put his life on hold waiting. He did not want to contact the mother, ever again. But he could have been contacted. 50 years later the child did their search.
    It was unsettling, but the child had the right to know and had been denied that right. Perhaps the father had the right to know though it would have been detrimental to his life at the time.
    But life is life. We have to deal with the results of what we took part in.

    If my relatives are rapist, or homicidal. Ouch. But do they not need to pay for the life they destroyed? Can they completely change and become all good? When you take what is not yours to take. There are consequences. Many pay in one way or another.
    Life is life.
    I still want to know who I am. Where I came from. What my history was. Where I am going with what tools and heritage.
    Thanks for the article Roberta. As always it was good, informative and to the point. Not a rage of spirit.

  31. I have collected DNA from me and my family för GENEALOGICAL purposes, nothing else. It is not ethical to use these results for other purposes whatsoever. If the states accept using the results, the there should be a found I can request the compensation for testing, as well as some licence fee for using my DNA.

    The tests are currently used not for their initial purpose and in an unethical way. I have halted all DNA tests and all uploads to GEDmatch. I have also removed my family trees from several sites, GEDmatch was the first one. I ma thinking about to ask FTDNA and AncestryDNA to remove all test results and to destroy the samples.

    Of course, searching for the murders and so on, but even I can sign up for this I can’t ask for a written consent all of my relatives when I upload my DNA. My decision is no DNA research anymore until the policies about the personal integrity are strengthened.

  32. Pingback: How DNA on an ancestry site identified the Golden State Killer - NewsBanc

  33. Someone mentioned above that a cup was taken after discarded by the (alleged) killer and the DNA then sequenced and uploaded to GEDmatch. Is that true? And if so how would they have done that? Wouldn’t it have been microscopic amounts of DNA on the cup? I presume the first sample they used from the rape kit was semen, so that makes sense, but I don’t get the second part of the matching story. Any info much appreciated!

  34. Pingback: 2018 – The Year of the Segment | DNAeXplained – Genetic Genealogy

  35. Pingback: Things That Need To Be Said: Victims, Murder, and Judgement | DNAeXplained – Genetic Genealogy

  36. Pingback: GedMatch Implements Required Opt-In for Law Enforcement Matching | DNAeXplained – Genetic Genealogy

  37. Pingback: Identifying Unknown Parents and Individuals Using DNA Matching | DNAeXplained – Genetic Genealogy

  38. Pingback: 2019: The Year and Decade of Change | DNAeXplained – Genetic Genealogy

Leave a Reply