Criminology - Genealogy Meets Science

Episode Date: October 6, 2018

We continue to explore the new DNA and forensic genealogy technics that are being used to solve cold cases. We're laying the groundwork, giving you all of the information you need ahead of the cases we'll be covering this season. In this episode, we talk to Colleen Fitzpatrick who has an extensive resume in using DNA and forensic genealogy to help give police names to aid in their investigation. We also talk to Steve Armentrout, the CEO of Parabon Nanolabs, whose company has been instrumental in many of the recent cases that have been solved using DNA and forensic genealogy. You can help support the show by going to patreon.com/criminology An Emash Digital production Learn more about your ad choices. Visit podcastchoices.com/adchoices

Criminology is a true crime podcast that may contain discussion about violent or disturbing topics. Listener discretion is advised. I'd like to welcome everyone back to episode two of season four of criminology. We're excited to get into this episode. We had a lot of people reach out to us after episode one and tell us how much information they took away from that episode and how interesting it was. And we thought that episode would be informative and in depth. and we're glad that people found it to be. So. And this episode is going to pick up right where we left off last week.

And we have some more great guests who are going to offer some terrific insights for the cases this season. But before we get started, we have to thank our Patreon supporters. Stephanie Levin Dusky jumped out at our highest level. Alicia Kirkpatrick, Michelle Miller, Angela Rogers, Joe DeRowell. and Cat Steel. So big shout out to all of the new supporters. And big shout out to everyone that continues to support the show month after month really goes a long way towards helping Morph and I put out this podcast.

Yeah, we do appreciate your support. And we've been trying to throw out some goodies out there on the Patreon feed for you to listen to, whether it's interviews in their entirety, outtakes or stuff we couldn't get in this show. or even fun stuff like the Q&A episode we just put out. If you'd like to support the show via Patreon and get access to some of those Patreon episodes, you can do so by going to patreon.com slash criminology. All right, Morf, let's start off this episode by recapping where we left off in episode one. We heard from Paul Holes, who gave us some more great insight about how he used emerging technology

to take down the Golden State Killer by dropping the killer's DNA profile into Jedmatch. Then we talked to Curtis Rogers from Jedmatch, and he filled us in on how Jedmatch worked. But with as much information as we got in that episode, there's still a lot more that we need to digest to really dive into and understand these cases that are coming up. Great police work is just one step of that process. and Jedmatch may hold the answers that investigators need to crack some of these cases, but there's a crucial step in between, and that's the forensic genealogy work. The investigators in these cases aren't qualified to do that work, and they rely on the help of skilled researchers to complete that work. In the last episode, Paul Holes told us how Barbara

Ray Venter connected some of the dots that helped him solve the Golden State Killer case. We've also seen how people like C.C. Moore, who has head of Parabon's genetic genealogy unit, is achieving similar success. One more person who's also been in the news often this year for her work on these kinds of cases is Colleen Fitzpatrick. And she has quite a resume for her work on decades old cases, including the case of missing aviator Amelia Earhart. She's also an accomplished author. She joined us to talk about some of her work as it relates to using DNA in identifying people, whether it's a serial killer or an unidentified body. She also gives us a behind the scenes look at how she played a role in solving some big cases recently. Colleen, thanks for joining us today on Criminology.

My pleasure, Mike. Glad to hear from you. If you can, please give the listeners a little bit of a background. about the work you do and what kind of experience you have in this field? Well, you know, Mike, I have a hard science background in physics, and I did a lot of work. That involves, you know, a lot of data analysis, a lot of statistics, you know, a lot of, you know, handling a lot of, well, a lot of data. And so, you know, dealing with the genetic genealogy cases that we have through the DNA Doe project and through identifiers,

it's been a natural for me to become involved and be successful because, you know, I know how to look at data. I know how to analyze it. I know how to tease out some meaning from that. So people ask me a lot, how did you get from science and physics and the forensics? And that's the answer. They're actually closely related.

It's how you, you know, work on solving a problem and using the scientific method. And so that was sort of, you know, basically my transition there was not that difficult. I know your name has been in some different news reports as of late, obviously with all the new advances that are going on with law enforcement and genealogy and the science behind that. How and when did you make that change going into this kind of work that you're doing now, this specialized kind of work? Well, about 2005, you know, I was working, I had my own company with a partner. We were working for NASA in the Department of Defense designing high-resolution laser measurement and fiber optic measurement techniques.

Our last contract was to help North Earth Grumman, a large aerospace company. We had a subcontract with them to help design an ex-spacecraft to Jupiter. It was called the Jupiter Icy Moon Orbiter Contract. spacecraft. And our team, we had a contract to help them produce a conceptual design, and there was going to be a

down-select because there were other companies competing for the actual construction phase. And our team won, which meant really a big deal for our company as a subcontractor. But unfortunately, right after that, the contract was withdrawn because, you know,

the Bush administration wanted to go to the moon as a way station to Mars. And regardless of that decision and why it was made, it meant our big contract went away. So at that time, my partner and I realized that it wasn't an era for small businesses. You know, there was another plan agenda, you know, for larger companies and for other priorities. So we decided to close the company and try and do something else. and at the time I had my first book, Forensic Genealogy, done.

And unfortunately, the publisher canceled the contract because it had taken me three years to write it. So with the book done, my partner suggested we just do it on our own, that at that time self-publishing was not what it means today, where you upload to the Internet and get an e-book or something. We actually opened our own publishing company, and we researched the printers, the paper, the ink, how much it cost, you know, even the format of the book, how many words on the page, you know, how to, and we went through it again and edited it very well. We're both very good writers. And so we managed to do this book, and we decided to use forensic genealogy sort of as a transition to whatever we did next, you know, thinking it would be more laser work or more optics work, maybe consulting.

that area, but it turned out the forensic genealogy idea took off and started a revolution in genealogy. So that's how we got started in that. And as of late, you've had some really good success. You've been, you know, contracted to do a lot of projects involving identifying people. What are some of the high-profile cases that you've worked on in recent times? Well, you know, that's a good question. You know, in 2005, okay, this developed outside the hobby arena. And I have done some really good work starting at the beginning from, you know, some of the cases I'm well known for, the hand in the snow, Titanic Baby Amelia Earhart, Abraham Lincoln, you know, that kind of thing, Holocaust work. But about 2011, I started working cold cases with the police.

And at that time, the only ability to do that using genetic genealogy was with YDNA. So after, you know, doing a few cases, I finally connected with the Phoenix Police Department and I wound up solving the Phoenix Canal murders. That was my first big whole case. I actually solved it in 2014, but the agency asked me to keep quiet about it because you can imagine this was the first case, saw it with genetic genealogy, and nobody knew what to do about it, you know, what it meant.

And I used only public data. You know, I didn't, you know, log into any secret databases. I just knew how to harvest that data from the Internet and use it to compare their wide DNA profile to the killers and come up with a name with that. So they finally arrested Brian Patrick Miller. I had given them the name Miller as the, the last name of the killer as probable last name of the killer and that narrowed their

list of suspects from 2005 and of those five there was just one that made any sense and that was the man that's now going to be tried for those murders because of course they followed up with you know getting a DNA sample from him and getting a CODIS match so it wasn't just I told him something and they went out and arrested him they had a lot of police work to do after I gave them that name. So I provided the final clue that led to the arrest. Now, more recently, fast forward, I've been involved in many cases since then, but about earlier this year, there was an arrest in a case from 1984 in Aurora, Colorado. In that case, there was a series of break-ins, robberies, murders, rates, of four residents.

in the Aurora, Colorado area. It's called the Bennett case because one of the break-ins was with the Bennett family. The four members of family, mom, dad, and the two girls were bludgeoned and raped and stabbed with only one survivor, which was the youngest daughter, who I think was four years old. Her face was smashed in, but she survived. So, you know, fast forward. You know, they hired me a couple of years ago to look at the case. And I came up with the name Ewing.

And this was interesting because it was probably the best match I had ever found. I found 75 matches in the genetic genealogy databases to men with the name Ewing to their profile of the unknown that they gave me. And this included exact matches, one step, two step. It was as if this was a huge old family of Ewing's and he fit right in. Okay. So, you know, barring the possibility he was adopted or, you know, he had a name change. We call a misattributed paternity or something. His name was Ewing.

The authorities, you know, and I, with my assistance, we looked at Ewington area, you know, that type of thing. And nothing came of it. And finally, about two months ago, they arrested Alexander Christopher Ewing for the murder. And the story was that once he had confirmed. committed these break-ins in Aurora, he moved to Phoenix and he did it again, but he got caught and he went into prison. And my understanding is that the prison he was in in Arizona probably was not adequate for him, not adequate security. Let me go back and explain that when all the original crimes happened in 84, 85, that was before Codas was enacted. So, you know, anybody who

was already in prison in the early 90s, when Cotis was. beginning to be used was not tested for CODIS. So a lot of the prison population has never been tested. So he was being transported to Nevada and to a more secure prison. And he got away, but he was recaptured. And then the state of Nevada recently has been going back to retest the prison population, having gotten CODIS hits off of inmates who were already in prison when CODIS was enacted. COTUS was enacted, and theoretically, everybody in prison and going into prison should have been tested.

But the funding necessary to test everybody already in prison was not there. So the decision was made to just test people going into prison. And that's why Mr. Ewing went under the radar for so long. Finally, the state of Nevada realized what was going on and started to test the prisoners they already had in prison. And that's when he came up as a Codis match to the Aurora. case, the Bennett murders in the Bennett case in Aurora, Colorado. That's a long way around of saying that that was one of my success stories, although you may not hear my name associated with it because once this was done, you know,

there's other issues that are being addressed in the media about retesting or testing the prison populations that have not been tested. You know, the, now the casework that has to be done to put. Mr. Ewing in the right time and right place, you know, the time, timeline and stuff. There's a lot of work that has to be done to prove out the case. So that all being said, that's more in the news and the fact I supplied the name two years ago. And just to follow up on that, I had heard what you just said, that there was a match in

CODIS, but they didn't really give you much credit that I saw. When you gave them this name, first off, did they check him out, or did you just give them the last name and they didn't, you know, connected to him right away? Well, they didn't even know anything about him. I gave him the last name, and we looked at Ewing's that were readily available. You know, Ewing that lived in Aurora at that time, you know, the police probably looked up old addresses, you know, traced people. But Mr. Christopher Ewing, you know, was nowhere on the radar screen. You know, he was already in prison in Nevada. So they didn't think to go and look in the prisons, you know, for somebody.

who was already in prison. They were, you know, looking at people not in prison. They thought he might have died. You know, they didn't know. That's fascinating. And I'm glad you brought that up because that's something that I wasn't aware of. So what difference is there, if any, between trying to ID somebody like Nichols slash Chandler or the Aurora Hammer Killer or a body, a John Doe, a guy,

Jane Doe who's found someplace that needs to be identified. Is there any difference on your end approaching it and investigating it to come to a conclusion? The difference between, say, let me say a John Doe, which would include Mr. Chandler, because he was a John Doe and a serial killer, it's not so much on our end what we do and how we work. It's more likely on the family's end because you're really walking into or you're helping with a situation, different situations. In one case, let's say you have a serial killer case, you're working with some police agency, and you have a bunch of victims, a bunch of families. It's in the family's best interest to keep that case in the news, to keep the authorities on track, to let people know

it hasn't been solved yet. And when you do the genealogy work, let's say you came up with somebody named John Smith that solved it, you know, we all always go to the agency, we'll let the agency handle it, but the family is going to be more than relieved. You know, the family is going to be, you know, grateful. A lot of them, you know, go on the news to say we're just so relieved. We know who my daughter's killer is, and we hope to bring him the justice. But on the other hand, when you, when it's John or Jane Doe, when you identify them, the family that the agency is going to contact has no idea that you're getting ready to contact them. You know, they, they have wondered what happened to their family member for years,

and they thought everybody forgot. You know, there's a few cases, say, that are very interesting, and the name, orange socks, lavender dough, buckskin girl. But for the most part, you know, the family hasn't probably heard too much. There hasn't been too much movement on the case. And when the authorities, you know, contact them, it's a surprise. And of course, they're, they are, in a different position. You know, it's hard, it's easy to say, hey, we found your daughter's killer. It's hard to say,

we found your daughter, she was strangled 37 years ago. That's a shock. You know, or your son hung himself in a closet in a hotel 17 years ago. That's a shock. And those people are, you know,

have been hoping their family member would come home. You know, they don't expect the news they died so long ago and they've been unidentified. And those people, you know, need time. They need space. They're not on the news. They don't want to necessarily go in front of the camera and say, hey, I'm so glad they, you know, identified my daughter. She was laying in a ditch

37 years ago. They don't, you know, they need time to cope. And so we have to, you know, doing the two types of cases, you have to be cognizant, not of what you're doing, but what you're going to find on the other end. You, I guess you have to somewhat have a feeling of, accomplishment or happiness for helping family members get news, even if it's painful news, you get some kind of good feeling from that, I assume? Well, when we saw a case, there's nothing like it. I mean, it's Miller time, you know, to quote, you know, Margaret and I, when we did Chandler that night, I mean, in all of our volunteers, the whole group was on fire.

You know, we were just so high-fiving each other over the Internet on the phone. there's really nothing like really making it work and bringing satisfaction or bringing information to the family. And I don't want to use the word closure. We use it on our end because we would close a case. But we don't want to use it on the other end because like buckskin girls' mother told us, there is no closure. You know, her daughter is never coming back. There's only information.

There's only knowledge of what happened that she can now go on with her life and cope with. You know, so there's accomplishment. There's also a sense of responsibility to continue to help, to be available as much as we can, to navigate the next step, you know, when the families have to do whatever legalities, they have to take care of now that they know the truth. As these new cases are being solved and more and more likely to be solved, and this seems to be a valuable field that you find yourself in, is there room for more people to come into that field? You know, because I've seen on social media people that are like, wow, I really, how interesting, how great a job that sounds like, is there room for other people to come into that forensic genealogy field and try and help solve these crimes? And how do they go about getting into that field?

I would say that there's plenty room. There's plenty cases. There's plenty to do. Now, in that regard, I want to just tell the audience that to do this, you have to have certain. skills. Of course, you have to have genetic genealogy experience. You've got to know, you have to have worked on, say, adoption cases or, you know, your own genealogy. You have to have some

experience behind you. But don't think that that's enough necessarily. You have to also know how to deal with law enforcement. You have to know how, you have to be very knowledgeable on how to approach people about

it. We don't call anybody, necessarily unless we really have to with the permission of the agency. But you've got to have people skills. You've got to have experience talking to people about sensitive subjects. You've got to have experience working with law enforcement. So I'm going to say to the audience, yes, there's plenty of room, but be very judicious about whether you want to hang your shingle out and do it on your own. Because it's a very volatile time.

Law enforcement is just learning. the ropes, they're learning what's going on. And, you know, we need to have the experience people speak with them and participate with them and, you know, work with them rather than a genealogist seeing an opportunity for, you know, to earn money or to open their own company doing this. You know, I, until recently and even recently, I'm one of the very, I would say not so many genealogists can, the police will return their calls or will. trust them with sensitive information. You know, I think I'm one of those people because I've

worked with them for so long and I have the reputation that I'll just be quiet, keep my mouth shut and work with them as they want. And I'm not out for publicity to make tons of money or to make a name for myself. I'm just out to do good work and help, and I'm well known for that. And how long is it typically a turnaround time if law enforcement gives you a case today and says, I need your help with this. Is there an average time until you can comfortably say I think I know the answer here? You know, the answer, I get asked that every single day. And the answer is anywhere from four hours until never.

And I don't want to say never, never, but four hours to a very long time. We have Bugskin Girl got solved with four o'clock hours. We had probably Margaret I, and let's say, other genealogist, it was like one to four in the morning, you know, so it was like maybe 10 to 15 man hours did that with Margaret and I, maybe a genealogist or two online and offline. Chandler took 3,000 hours over several months, and we have two cases that we're still working on. We're actually making progress finally. One, the problem was the DNA was contaminated with bacteria. So that meant we had a lot less to work with. And then we

anticipated. The other one comes from West Virginia, Bell in the Well, and her problem is she's so from an area that's so intermarried that it's hard to untangle the genealogy from, you know, who's who, who's related to whom. And we also found that that's a strike against finding one parent, but the strike against the other parent probably is that the other parent probably came from Germany or continental Europe. And there aren't that many people. in the database, relatively speaking, that are from those areas, and this particular person, whoever the parent is, really doesn't have any close matches. So we're at a disadvantage on both sides of their family. And at any given time, how many different cases are you working on,

or is there a limit to the amount of cases that you can be working on at once? Well, we're ramping up. We have 40, 45 volunteers. We have a waiting list about twice that long. And we're ramping up, and we have probably 20 cases or two dozen cases anywhere from, you know, very mature and almost solved all the way to just coming in the pipeline like in the next couple of days. So, you know, it's a, you know, it's not just we're working and we're not working. We're working or we're pulling them into the system. So we think we could handle a lot more. And as I said, as we grow, we don't want to grow too fast. We don't want to have too many volunteers, which means everybody's working on that next new case.

And we don't want to have too few, which means some language in the corner. So as we get more cases, we're going to be adding more people. Well, if your past work and history of accomplishment is anything, you'll probably have plenty of work coming your way, and hopefully you solve a lot of it, and I want to wish you luck with that. Yeah, thank you so much. The Bureau of Justice Statistics survey of 2007 said there were 40,000 unidentified bodies in various, you know, custody of various agencies and medical examiners office.

Sometimes some people regard that as a vast underestimate. Some are okay with it. But either way, the number in 2018 is so much higher, and we're out to identify all of them. And if somebody out there listening has a need for your services or just wants to learn more about what it is you do, where can they find you at? Well, I'm Identifenders International as my company. My blog is Identifenders.orgpress.com. My website is www.W. Identifinders.com.

And that would pertain the more criminal cases or, you know, where I could apply YDNA, if there was no other DNA, still available. and, you know, other cases, you know, some heavy adoption cases are, you know, Holocaust. I do a lot of Holocaust work. As far as DNA Doe, you can go to www.DNA Doe Project.org. And, you know, there's a contact us there. We'd love, we survive on donations, so we'd love it if people would surf in and donate to their favorite cases or just to our general fund to help us keep going. We're totally nonprofit.

not draw any salaries, no people make any money off of this. Any administrative fees we earn, go to, you know, all of our expenses to support our website, our attorney, you know, our CPA and that kind of thing. So go to the, and if you'd like to suggest the case, we have, of course, tons that have been suggested. One thing we don't do, if somebody writes in, said, could you do this case from Alabama? We don't go and contact the agency on the suggestion of a right-eat-exam. in. We suggest that the agency contact us and we'll consider it. We don't encourage genealogists to call their agencies to suggest cases, but, you know, I guess email doesn't hurt or whatever,

but mainly we get our cases when the agency comes to us, not vice versa. And are you on social media at all, Facebook or Twitter? Do you have any pages there? Yes, we are on social media. We have a Facebook page. Search on DNA Doe project and you'll find us. Awesome. And we'll be sure to link to some of your information on our site or on social media so people can find you as well. I appreciate the information. You've been a wealth of knowledge and you're really helping us to understand the stuff we're talking about this season on criminology. So, thanks for coming on.

Yeah, thanks, Mike. And we'll talk soon. Colleen mentioned having a hand and identifying the Aurora Hammer killer Alexander Christopher Ewing. But she hasn't really gotten a lot of credit for that because ultimately DNA on file and CODIS matched Ewing. But that was such a well-known and horrific case. So for her to play a role in that is just amazing. So in that interview, you heard Morph asked Colleen about Robert Ivan Nichols,

aka Joseph Newton Chandler III. And that's because Colleen also worked on that case. help to break it, and this is one of the cases that we're going to cover later on this season. You'll hear a lot more detailed information from Colleen in that episode about the success that she had with working the Nicholas slash Chandler case, which is really fascinating. And you heard Colleen allude to a mention of somebody taking their life in a hotel while using a fake identity. And that was the case of Lyle Stevik, which she also had a hand in solving. And of course, they haven't released his true identity, but she did accurately figure out who he actually was. And I know a lot of listeners out there from what she mentioned probably would figure that was Lyle Stevik she was talking about.

So far this season, we've talked about Jedmatch. We've talked about the genealogy used to identify some of these people. but one final thing we need to jump into is the use of Parabon nanolabs in assisting law enforcement over the years, and they've been around for a while. The work that they have done and continue to do is impressive. And Parabon is another one of those names that we see time and time again in relation to some of these solved cases. I think many people may be familiar with Parabon, especially in relation to the ground. breaking work they've done in creating composites of what a person may look like based on their

DNA. Parabon did play a major role in many of the cases that we've seen solved this year, and some of those cases we'll be covering this season. Steve Armantrout, who's the CEO of Parabon Nanolabs, joined us to give us an in-depth understanding of just who Parabon is and what they do. With us today is Steve Armantrout. He's with Parabon. Thank you for being with us, Steve. Of course. If you can, just introduce yourself to the audience about the work you've done with Parabon and your background.

So I'm the CEO here. Parabon Nanolabs was founded in 2008. This is a subsidiary area of a company called Parabon Computation that I've ran for a long time. And Parabon Nanolabs is a DNA technology company. We both build things out of DNA and we analyze DNA. And so for the forensics world, we analyze crime scene DNA in a different way than traditional methods. And when you first came online, did you anticipate this was the work you were going to be doing? A lot of it specifically centered around law enforcement?

Absolutely not. We first started this part of the business doing DNA nanotechnology, and it was only thanks to a government solicitation looking for something called DNA phenotyping that we got into the business. The company has a long history of working on difficult computational problems, and in 2011, the Department of Defense issued a solicitation looking for tools that can predict appearance from DNA. And the way it was couched, it was sort of like they were looking for a lab kit. But when we looked at the problem, it was really a computational problem. Can one reverse engineer the human genome to figure out what genes are responsible for eye color, hair color, and so on? So we thought we had a unique angle on this problem. We responded to that solicitation with a proposal.

And of all the companies that submitted, we got the award. That was a phase one project. The goal was to establish the feasibility, answer the question, can you find the genes or the SNPs, as we say in the business, the single nucleotide polymorphisms, the changes in the DNA that are responsible for various traits? And we were successful in that endeavor that led to a phase two project. And what evolved over the course of about a four-year period of time was a technology that's now called snapshot DNA phenotyping and snapshot DNA analysis. looks at the DNA differently than traditional forensic analysis. The forensic tools with DNA to date are fabulous for what they do, but they treat the DNA like a fingerprint, like a biometric, like an iris scan.

And so they're very good for that identical matching, identity matching, but they really don't take advantage of all the genetic information that's available in DNA. So with Snapshot, we use different instrumentation and we look at that genetic component. And over the course of the history of that snapshot development project, we developed models, predicted models for predicting eye color and hair color, skin color, freckling, face shape, and ancestry. And in 2015, we began offering that to law enforcement as a service offering. And I think that's when a lot of people first heard about the work you were doing with these these almost composites that you were producing based on what the DNA was telling you. And that seemed like a revolutionary phase in DNA work.

What were the limitations of working with those, you know, as far as accuracy and dependability? When we first got into it, there was an open question about whether you could even analyze forensic-grade samples with the instruments that were. we use. We use what are called SNIP microarray sequencers. And the SNIP microarrays typically require a lot of DNA more than you find in a typical crime scene sample. So we and others analyzed whether that could work. Once we demonstrated the instruments were indeed sensitive down to levels as little as one nanogram, an extraordinarily small amount of DNA. Then the next

question was can you build models that are predictive? And for that purpose, we collected DNA and phenotype, i.e. trait information, on thousands of people. And using them as references, we developed what are called now AI, but machine learning algorithms to make predictions based on these examples. The core idea there is if you have enough examples of DNA and blue-eyed people and DNA of brown-eyed people, maybe you can find genes that turn on or off depending on whether it's brown or blue. And I'm oversimplifying, but that's the basic idea. So as part of this process, we keep track of how well the models predict on our known samples. So before we ever went to market, we had very good data on the accuracy of the various models,

because we build them in a fashion that allows us to test on a part of our sample. It's kind of a clever idea. Briefly stated, it's called cross-validation, but you can imagine taking one of your samples out that you know about. You know the eye color of this person. And I'm using eye color because it's a simple phenotype to talk about. You know the eye color of this person, but you treat their data as an unknown. You don't let your models see that. And so you train up your models as though you never had that sample.

and then you can analyze that sample with the models you've built and see how well it performs. And you can record that. Then you can put that sample back in your dataset, take another sample out, do the whole process again. And over time, you create a log of just how well a model performs. And what we found, through our delight, is that the models are very accurate. And we actually report our confidence in those predictions along with all the reports that we produce. And are you able to give like a percentage of how accurate you believe these composites to be? So for the various pigmentation traits, we provide these confidence statements.

And, you know, for eye color, hair color, those kind of things, we're in the 90% confidence range. We've done a lot of blind evaluations and so forth where that seems to be the case. And every so often we'll say, you know, green eyes or hazel make those kind of subtle mistakes. but overall those models are very strong. Face morphology is more difficult to measure. But in the end, in our database, we're able to discriminate three out of four faces just using face morphology alone.

So that gives you some indication that these are pretty strong. And in the end, what we're producing is really a likeness of someone. And that's proving to be useful for law enforcement. From a law enforcement perspective, their primary goal starting off is just to exclude who can they de-emphasize in an investigation. And so we don't claim that the composites are photo IDs.

In fact, we have a briefing with our detectives on each case, and we underscore the point that this composite is intended to be a likeness of the person whose DNA came from, not a photo ID. And so they use these markers, these phenotypes, in their investigation, to narrow a suspect list or to jog someone's memory. And in that capacity, they're proving very effective. As far as technology goes, are these still on the newer side of what's evolving in DNA technology? Or are there something new on the horizon that might take the place of these composites? This is very much the cutting edge of what's going on in DNA forensics today.

There's phenotyping, which has been around for a couple of years now. We introduced a test for kinship inference. So this is the ability to take two DNA samples, and these are arbitrary unknown samples, and say with high confidence the degree of relatedness between them, if any. That was part of a Department of Defense contract. where it's ongoing, where we assist them with the identification of war dead. You can imagine in a past conflict, say even back to World War II, there may not be a next-of-kin, and so you're not able to use those traditional DNA tests, whereas this test can go out to,

you know, second cousins once removed with high accuracy. And then most recently, we've been using something called genetic genealogy, and that's what has made the news recently. CC Moore is a well-known genetic genealogist who joined our staff in May. And that's using the DNA along with a public genetic genealogy database called Jedmatch to try to come up with relative, genetic relatives of an unknown DNA sample. And that's proving to be highly effective. In fact, since May, Parabon has been part of investigations

that have resulted in a positive DNA match 11 times now. And some of these cases are over 35 years old. And congratulations, because I know just in the last day or two, there's been a couple more that you guys have hit on. And I know you can't necessarily speak about some of the cases that you are working on for law enforcement. But in general, how many cases do you have, as far as a caseload of cases that law enforcement is asking you to review?

You know, there's various stages in our pipeline, but if you go end to end now, we're probably looking at 50 to 75 active cases and more coming in every day. This technology has marked a real turning point in the history of DNA forensics, and the law enforcement community is excited about it. And do you limit yourself on how many cases you want to work on or you can work on at one time? We're growing to match the demand. we brought on a number of genetic genealogists as the case load has ramped up. And so it's just a matter of managing that growth. And what's the process from, you know, if you can give me a brief overview, from law enforcement contacting you saying we have this DNA from this offender?

Or in another instance, you know, we've got an unidentified victim that we found out in the woods. We don't know who it is. from getting that DNA, however, they're giving it to you. What's the process, the turnaround time from the start to the finish when you have a name? So the process, many steps, I'll summarize them briefly. First, there's just a screening process. In other words, we can't accept just any case. There's got to be DNA and there's got to be sufficient quantity and it has to be of sufficient quality.

So we have screening forms that the DNA analyst for a particular case, will fill out and then our team can evaluate those cases and we'll either you know accept a case or turn around and say hmm you know this particular sample is high risk can you get another extract is there other evidence that you can test and so on in some cases you know we just have to say no we can't take this case because this is the last of their evidence and the DNA they have isn't isn't sufficient but once we've accepted a case then we can accept a sample. We'll have that sample shipped to one of our genotyping labs.

They'll run it through the instruments. And what comes out the other side is a big text file. You can think of it as a spreadsheet with a million rows. And these are the alleles, the genotype calls, for the various stint markers. So with that genetic information, we then start applying our tools. We do a number of things. We'll do a genetic. genealogy screening. So that involves uploading the file to Jedmatch and recording the top matches and how much DNA is shared. Because if we don't get matches that are sufficiently close, we really can't do genetic genealogy. So that's sort of stage one. Additionally, we'll do basic phenotyping. So we'll run through models that predict ancestry,

eye color, hair color, skin color, just the basics, because it turns out those tools are helpful when our genetic genealogists go to do their work. And then if the police want to proceed, if we've gotten good enough matches to warrant a deep investigation, they'll give us approval and our genetic genealogy team will then go off and begin doing their detective work. And quite frankly, that's what it is. They're using DNA sharing information that comes from jet match to identify matches. Then they do genealogy work. They've got to build backward into the family trees to find the most recent common ancestors of the unknown and a match. If we're lucky we get matches in different parts of the overall family tree, so you can begin to

triangulate trees from one family to another. And ultimately, you're looking for that marriage that led to an unknown. So that work, you know, very labor-intensive. We have some of the best genetic geologists in the world, and they will take our phenotype information. Sometimes we'll do targeted kinship testing, maybe a volunteer in the family, who will give a bocholab.

And with that, we can either include or exclude large branches of the tree. But at the end of this process, will provide a report to police that gives them, you know, in the best case, here's an individual's name. It might be a family. And in really hard cases, it may be, you know, here's a list of surnames that you should be looking for in your suspect list. Here are the most recent common ancestors, but there may be lots of descendants. So at that point, there may be a collaboration that goes on where the police. have to do a lot of good police work, and they'll come back to us.

Here's what we found. Here's a sample we want to have tested, and together we'll work on the case. And what we're seeing is these kind of leads are providing police with great starting points for these investigations, and they're finding positive DNA matches, you know, with increasing frequency. So the timelines that it takes the lengths of time it takes to work on some of these cases can vary from one case to the other. Sure. It all depends on the samples. It depends on how many matches one gets, how close they are. And, you know, if we get lucky and we have close matches, we can turn them around in a matter of a couple of days.

And then a more typical turnaround is 45 to 60 days. And what do you see next is the future, or would you predict, is the future for Parabon and the next phase of work that you'll be doing? Well, we're introducing a new software suite called Parabon FX that we're excited about. It's really a collection of algorithms to analyze next-gen sequencing data. So even though we're doing genetics work on samples today, it's using a technology, as I described earlier, called SNIP microarrays. Next-gen sequencing is a different way to analyze DNA. And it's not going to replace SNP microarrays, but at the same time, it's applicable to tougher samples. It's a little better for analyzing mixtures.

This is where you have maybe. multiple people have handled a gun, and you want to be able to tease out those profiles. So next-gen sequencing is a tool that you're going to see being used more and more in forensics, and we're on top of that. This was a DOD-funded effort that were installed in a platform that, in fact, next week at the issue conference, we'll be announcing it. And it seems like as if the public is generally supportive of all the work you're doing and other agencies, you know, similar to you doing this work that's catching a lot of these bad guys that we're seeing in the news.

But there's also people that have privacy concerns or concerns about errors or misuse, mishandling, perhaps at the laboratory level. What can you say to alleviate any of those concerns? Well, the first thing I would point out is that, you know, we're providing law enforcement with leads. And by definition, any case that we work on, they're going to have DNA on. hand. So at the end of an investigation, they're ultimately going to use traditional DNA matching technologies to prove that a suspect is in fact a match. So that's just an important fact to understand. Our tools are not going to be used for that final claim that this is the match to the DNA because they're existing tools and protocols in place. With respect to privacy,

I think right after the announcement of the Golden State Killer case, there were some privacy concerns raised. But as folks really began to look closely at the actual practice of forensic genetic genealogy, you begin to realize that the practice isn't imposing on anyone's privacy. We're using JetMatch, which has as part of its terms and service, it states that law enforcement may use this website. So if you're a Jedmatch user, you have the option of making your kit private. So you can still use the tools. It's just no one can see your data. So you have complete control over that. And moreover, the data on Jedmatch is really not very sensitive data.

It's not raw genetic content. Jedmatch simply calculates the amount of DNA sharing between other people in the database and your file. So when people really begin to examine the actual use of the tools, you realize that no one's privacy is being violated. And I think that alleviates a lot of concerns for a lot of people. But overwhelmingly, I think that people are supportive and very positive about what you're doing and what law enforcement is doing with the DNA. We certainly seen that. It's quite moving to get responses from the families and from people in the communities where these crimes have been committed, thanking us for our work. So, at least from our perspective, the response has been positive.

So going forward, obviously you've done a lot of good work and you continue to do a lot of good work. What are the chances one day that Parabon, as a business, could sell its work or the business itself to a private company? ancestry.com, for example, and then that company would turn around and reverse the work you're doing now, we're no longer willingly assist law enforcement. That seems pretty unlikely. I mean, particularly the snapshot brand and the division of people that support this process, we wouldn't be very attractive to a suitor as anything other than a forensics business. So that seems unlikely.

Moreover, you know, other people are. beginning to pursue these technologies too. So I think that risk is pretty small. Well, I want to wish you continued success because so far you guys have proven to do outstanding work. And I think is that weeks go along and we see more arrest and the name Parabon popping up. It's just a real good positive thing you guys are doing. So more power to you and good luck moving forward.

Well, thank you so much. real pleasure to be on your show. I appreciate it coming on, Steve. That was some great information that Steve shared with us, and we can't thank him enough. I think when we combine everything we know about Parabon with what Colleen Fitzpatrick shared with us, as well as the insights from Curtis Rogers from Jedmatch and Paul Holes, we really get a much clearer picture of how the process works from beginning to end. Whether DNA is being used to help ID the family tree,

of a serial killer or an unidentified body. So now that we're armed with all of this information and have a better understanding of how these processes work together, I think that we can dive into one of the cases solved in 2018 using these tools. And that's the case of Jay Cook and Tanya Van Cullen Board, who were murdered in the state of Washington in 1987. But that dive into that case will come on the next episode of criminology. If you like the show, please take a minute, go out, give us a five-star rating, make sure you tell your friends about it.

All of that goes a long way towards helping other people find the show. And if you'd like to find us on social media, we're on Twitter with the handle at Criminology Pod. You can also find us on Facebook by searching for Criminology Podcast. or by joining our Facebook discussion group, criminology podcast, discussion and fans. All right, everyone. Get geared up for next week, full-blown, in-depth episode

on one of these cold cases that was solved with the use of the technology that we've talked about in these first two episodes. To another round of drawing board or mirror board. Today we talk brainstorms with U.X designer Brian. Let's go. First question. You thought you'd see everyone's idea in the team brainstorm, but you've got a grand total of one. Drawing board or Mero board? Drawing board, right? Because in Mero, the team can add ideas now or later. And with privacy mode, we can keep them anonymous until they're good to share. Correct. Next, you need the best way to explain your idea, but all you have is a few sticky notes. Drawing board or Mero board?

Drawing board, because, you know, in Mero, I could record videos at text, images, links, and digital sticky notes, of course. thoughts the way I want. Right again. Now, you're looking for a past idea you thought was just genius. Only you could find... Oh, there it is. Drawing board or... Miro. All our finished and unfinished work lives in one place. And he's one. Join over 60 million people getting ideas noticed in Mero brainstorms. Get your first three boards for free at Miro.com. That's MIRO.com.