How to catch a killer with your own DNA

Science in Focus Series – How to catch a killer with your own DNA

Marie: Hi everyone and welcome to Science in Focus, a free public lecture series showcasing the latest research from prominent UTS researchers and scientists. I'm Dr. Marie Morelato and I'm a Senior Lecturer and the Course Director of the Bachelor of Forensic Science at UTS, and I will be the moderator for today's session. I'll be asking our speaker any questions that you have and asking them at the end of the talk. So today it is very exciting. We'll find out how our DNA might be used to catch a killer. So forensic genetic genealogy has the potential to find the genetic relatives of any perpetrator who leaves their DNA on a crime scene. So that means that anyone and everyone, including you, could be a genetic informant. So, to tell us a little bit more about this new development in forensic genomics, we are joined by two UTS researchers today. But before I introduce our speakers, I would like to acknowledge the Gadigal people of the Eora Nation upon which ancestral lands our city campus now stands. I would also like to pay respect to elders both past and present, acknowledging them as the traditional custodians of knowledge for this land.

Marie: So, before we start, I'm just going to do a little bit of housekeeping. If you have any questions during today's webinar, just please type them into the Q&A box in your Zoom control panel, and we'll do our best to answer them depending on the time. If you like a question someone else has asked and would like it answered in priority, please use the voting tool, which is the little thumbs up, that is the symbol that is next to the question itself.

Marie: So, this session will be recorded, but we will not be recording any video or audio input from the audience. So now to the interesting part of this webinar, our speakers. So we have two speakers today, Professor Dennis McNevin from the Centre for Forensic Science suggests, and Dr. Nathan Scudder, who is the Coordinator of Research and Innovation, Policy Development and Innovation at the Australian Federal Police. So, our first speaker is Dennis McNevin, who is a professor in the Centre for Forensic Science. His research has focused on all aspects of DNA profiling from the extraction of DNA from difficult substrates, like bone and hair, to the next generation DNA sequencing technologies, and the prediction of genetic ancestry from DNA. He has provided forensic genetic services to the New South Wales Police Force and the New South Wales Health Pathology Forensic Analytical Science Service and is currently seconded to the Australian Federal Police National DNA program for unidentified and Missing Persons. So, Dennis how does forensic genetic genealogy work?

Dennis: Thanks very much, Marie. I'm going to start off by showing you some results from genetic ancestry that I've received from what we call 'direct-to-consumer' genomics service providers. So, on screen now, you should be able to see a genetic ancestry result, which I've received from a company called 23 and me. And they have identified that most of my genomic ancestry is from North Western Europe, but in particular from the British Isles and Ireland with County Cork highlighted. So I've also submitted my DNA to another company called Ancestry DNA. And in both cases, what I had to do was to spit into a tube and send about 10mLs of my saliva in the post to these companies in America. And my second result here from Ancestry DNA shows more or less the same thing. You can see that my ancestry, my genetic ancestry has been narrowed down to the British Isles and in particular Ireland. Now, this coincides with research that my father did many years ago, 20 or 30 years ago, before any of this genetic ancestry was available. And he looked at birth records, death records, church records, and he was able to trace our ancestors back to County Cork in Ireland. So it's reassuring that the results I've received from genetic ancestry are consistent with the ancestry that my father was able to trace back through documentary records. So it seems like these ancestry predictions are more or less accurate, certainly in my case.

Dennis: Now there's lots of other information that you can obtain from these service providers. And one of the latest features from Ancestry DNA is that you can find out which part of your ancestry comes from either of your parents. So you can see here in this in this graphic from Ancestry DNA that we've got two halves to this genetic ancestry, Parent One and Parent Two. They can't tell which parent, so we don't know whether it's my biological father or my biological mother. But one parent has bequeathed to me mostly Irish ancestry, and the other parent has bequeathed Irish ancestry at about 50% and English North Western European ancestry. Now in these service providers like Ancestry DNA, you can construct a family tree that reflects these genetic ancestry results, and you can attach to these trees documentary records of the type that my father used to trace our ancestry back. So, what it means is that you can then construct a family tree that includes and incorporates both your genetic record and documentary evidence. And this is then used to locate other potential relatives that you might have. So occasionally I'll get a notification from a service provider like this that tells me that I've got a first or second cousin that may be included in my family tree, even though I didn't put that person there. And that's because they're able to predict, based on my DNA, that this person might be related to me and might be included in my family tree. And then I can go in and verify those connections.

Dennis: So how is this possible? How can these companies make these predictions and help me build out my family tree? Well, it's based on the genetic principles of inheritance and the fact that we share DNA with our genetic relatives. And the more closely related we are to somebody, the more of our DNA that we share. So, this graphic here demonstrates that principle where this is me in the middle, the self, and you can see that my close genetic relatives, my brother, my sisters, my parents and my children, they share half of my DNA by descent. Further out: uncles, grandparents, aunts, nieces and nephews share a quarter of my DNA by descent, and we can keep working our way outwards through these concentric circles to get to very distantly related genetic relatives like cousins once-, twice- and thrice removed. And of course, as we move further out, the smaller that that fraction of DNA shared with me.

Dennis: So, here's another diagram that shows essentially the same thing. This is in a family tree format. So here I am here in orange. Here's my parents. I share 50% of my DNA with my father, 25% with my grandfather, 12 and a half percent with my great grandfather, etc. And as we move over here to these distant relatives, like third cousin, twice removed, it's much less than 1% of my DNA that I that I share with them.

Dennis: Now, what's the physical mechanism of this dilution of DNA as we move away from my close genetic relatives? It occurs by a process called linkage. And this graphic here represents a family tree, or, as we call them, pedigrees. And down the bottom, we have two first cousins. Here they are here. They have two parents. And those parents also have parents. And they're linked by these common ancestors at the top here. So this these two individuals, the most recent common ancestors for the two first cousins at the bottom. Now, each of this first generation up here have got, where squares represent males and circles represent females, have got these two coloured bars and they represent chromosomes, which are the packages of DNA that we have inside the cells in our body. And so, these chromosomes, in the process of reproduction, they get shuffled by a genetic process called recombination. And so, the chromosomes that are passed down to the offspring are shuffled. And that's what's represented by these different colours in the bars in this first generation. And then when we get to the second generation, the first cousins, you can see that that shuffling has been even more, because it's shuffling upon shuffling. And so, it means that two genetic relatives will share stretches or fragments of DNA that are uninterrupted. And these are called IBD segments, IBD meaning identity by descent. So, this little section of DNA in orange here on the two chromosomes shared with the two first cousins, they are shared uninterrupted, and that's an indication that they are genetically related.

Dennis: Now here's a graphic which shows how that process of recombination and shuffling leads to a gradual disintegration of linkage blocks. So, these stretches of uninterrupted DNA will become shorter and shorter as we move through the generations so that eventually, as we move through many generations, 100, 1000, those little IBD segments become smaller and smaller. And so, the further you are away from a genetic relative, the smaller the shared blocks of DNA that you have. And those blocks can be measured, the lengths of those blocks can be measured. So, here's a table from the Shared CentiMorgan Project, where the lengths of these IBD segments that are shared with genetic relatives are indicated with possible ranges. Okay. So here I am in the middle, in the self. You can see that my parents should share about three and a half thousand centimorgans of DNA with me. Now, it's not exact. There's a range there, but it's about three and a half thousand. My daughter should also share about the same amount of DNA. My siblings should share about two and a half thousand centimorgans of DNA with me. And so on, we can move out through my family tree to more distant genetic relatives so that by the time we get to fifth cousin, three times removed, we're down to only about ten centimorgans. And that really represents about the limited detection of these genealogy methods.

Dennis: So, what does that mean? Well, it means that an unknown profile can potentially be identified using these genetic relationships. So, in this pedigree here, this black circle represents an unknown profile that we're trying to identify that has been uploaded to a genetic database. And some putative relatives have been identified because they have been suggested to be of the right genetic distance, to be equivalent to second cousins or first cousins twice removed. So, what that means is then we can then go to these potential genetic relatives, build up the family trees to most recent common ancestors, and then build the tree down again to the unknown profile to potentially identify that unknown individual. So that's the science behind genealogy. I'll now hand back to Marie, who's going to take us to the next section of the of the presentation.

Marie: Thank you, Dennis. So, let's hear from our second speaker, who is Dr. Nathan Scudder, who is the coordinator of research and Innovation within the Policy Development and Innovation Branch of the Australian Federal Police. He has worked in operations including the Bali bombings in 2002 and Indian Ocean tsunami in 2004. Building on his legal and forensic background, Nathan researched the privacy and legal implications of advanced DNA technology for forensic science and received his doctorate from the University of Canberra in 2020. Nathan is an Adjunct Associate Professor in Industry in the Centre for Forensic Science at UTS and is a member of the Australian Forensic Genetic Genealogy Collaboration, working to assess the feasibility of advanced DNA capabilities to solve crimes in Australia. So, Nathan, how did forensic genetic genealogy come about in law enforcement and how is it being used?

Nathan: Thanks Marie and thank you to UTS for the opportunity to present as part of the Science in Focus series. Before I start, just a few quick disclaimers. The federal government is in caretaker mode at the moment, and because I work for the AFP, to the extent any topics come up that are at issue in the federal election, I can only provide factual information. All other views and comments are my own and this presentation doesn’t constitute legal advice. I should also note the presentation I'm about to give does include a couple of case studies relating to violent crime.

So as Dennis has explained, that inheritance of DNA means that if two individuals are found to share segments of DNA, we can hypothesise that they have a common ancestor. The more DNA they share and particularly long uninterrupted segments of DNA, the closer that ancestor should be to the present day. So being able to identify individuals through these shared DNA is certainly of interest from a law enforcement perspective. But a number of things needed to happen in order to make this possibility a reality. So, in terms of the genetic testing that Dennis has explained, Monday was National DNA Day and you could buy one of these DNA test kits if you tried hard enough for about $40 or $50. Those prices were unheard of only a few years ago.

Nathan: When you're talking about crime scene evidence or samples from human remains, the price is still in the thousands of dollars, but that price has also reduced considerably over recent years. But law enforcement isn't going to pay thousands of dollars to sequence DNA if there's nothing to compare it to. And the size of the databases to which you can send your DNA have increased exponentially in recent years. As we'll discuss, law enforcement doesn't have access to all of those databases and those that do provide access, have an opt in, opt out arrangement for crime scene comparison. But we're still at the point now where there are enough profiles available. This technique has certainly been shown to be feasible in the US. And research here is showing that it is also feasible in the Australian population and many other countries. But even that's not enough. Because knowing that two people are second cousins doesn't help unless you can show exactly who their common ancestor was. When I first started researching my own family tree about 20, 30 years ago as well, if you hit a brick wall and that's a question you couldn't immediately solve, the solution was to go to the church records, to go to an archive, to check a cemetery. If you're researching your own family tree, that's feasible. If you can research these questions over many years, if you have money to travel, there are certainly some people who would argue you're not a true genealogist until you've mastered the microfiche readers at the National Library. But for law enforcement, that is not feasible. You need ready access to those records to build those trees as Dennis showed. And over the last 5 to 10 years, we've seen a huge digitisation of genealogy records in Europe, in the UK, in the USA and here in Australia there's a lot of information now available and searchable. In fact, only as recently as about a month ago, the US released the 1950 census. And an army of volunteers went to work to index all of those records so that genealogists could type in a name and find a census record, find out where their ancestors lived, what their occupation was, who was living with them, and so forth. So having that information at our fingertips makes this technique possible.

Nathan: So, I'm just going to walk through a couple of quick case studies. The first case was one of the very early cases where forensic genealogy was used. It's the case of the "Buck Skin Girl". Here's a quote: 'after decades of work, hundreds of leads it was to forensic genealogists at their computers in the middle of the night who cracked the case'.

Nathan: So, the Buck Skin Girl was murdered near Troy, Ohio, in the early eighties. She could not be identified. She was wearing a distinctive jacket, hence the name. But she was ultimately buried in a 'Jane Doe' grave. Investigators retained one vial of unrefrigerated blood. And that was kept for 37 years until the charity group called the DNA Doe Project raised money to sequence that blood. And I have to disclose here, I do some volunteer work with that organisation. This particular case well and truly predates that. They had that analysis done. What that analysis came back with, was genetic data with about half of the markers that you would get if you spat into a tube and sent it to one of these online providers. But that's plenty, that number of markers is sufficient. And by uploading that to a public database, which at the time had about 800,000 profiles in it, after the data had been analysed. The top of the list was a user who had uploaded their DNA. The amount of DNA they shared was that of around a first cousin once removed. So what does that mean? So in terms of constructing that family tree, we've got the match, we know that their parents and grandparents is about as far back as you then need to build. You're really looking at children of a sibling of the mother or the father. You don't know which. So you have to build down all of those lines. And then you start looking at all the first cousins and first cousin once removed. It's slightly simplified, but assuming you know the relative ages of the match and the Buck Skin Girl, you can conclude that the Buck Skin girl should be the child of one of those first cousins. And I said, you don't know which side of the family they're on. So you have to build down all of those lines to map out that tree. But you've probably only got a dozen or so individuals, perhaps, that you've had to map out from that information. Now, in the case of the Buckskin Girl, that match had information that led the DNA DOE Project team to a number of family trees on the Ancestry.com platform. And one of those close relatives had actually marked one of these individuals in the tree as missing, presumed deceased. Now, that's not enough. This is purely an intelligence lead. So you've now got to go back to the investigators, even with that enormous coincidence. They need to do additional testing. They need to find a close family member. Use all of our usual forensic processes. Take a DNA sample from them. Compare it directly to the DNA from the Buckskin Girl sample to confirm that that match was correct. They did all of that and they confirmed that the buckskin girl was Marcia King, detective quoted as saying, "We can start a victimology now", the tips have already started coming in. Now, this case is now four years old. It's still an open homicide investigation, but they have had some tips that have come through. So that was one of the very early cases, but only two weeks later, there was the case of the Golden State killer, and this is the case that attracted worldwide media attention.

Nathan: So the Golden State killer, also known as East Area Rapist, the Original Night Stalker, the Visalia Ransacker, the Diamond Knot Killer. This is a geographically mobile offender. He was committing burglaries, sexual assaults and murders up and down California in the seventies and eighties. He stopped offending in 1986. Now some have speculated, I haven't ever seen this confirmed, that that was around the time when DNA evidence first started coming in and the significance that will become apparent in a second, but that may have been why he stopped offending at that point in time. So this case was never really cold. They always had investigators working this case. And more recently, they started to look at big data analytics. So they were looking at municipal records. They were looking at prison records. They were trying to work out whether they could find someone who fit that description of the Golden State killer. And they had plenty of eyewitness descriptions. They had his DNA. But of course, it matched no one. But why was the Golden State killer in these locations at these times? Did he go to prison? What, did he die in 86? All these questions were being analysed. When they actually identified the Golden State killer and they went back to those lists, his name wasn't on any of them. So he would almost, almost certainly have gotten away with these crimes had it not been for this technique.

Nathan: So this time when they uploaded the DNA and fortunately, some of the investigators had retained additional samples from some of these crimes, that they could go back to and re-sequence using this this new DNA technique and get enough of those markers to upload to these databases. They didn't get a first cousin once removed. I understand their closest match was a third cousin. So we're talking about far more distant relatives. And they effectively took the top 25 matches and they started building back the trees. Now, this was this is quite an involved process. It took them three or four months. They had investigators going to cemeteries to check dates of death and so forth. It was quite an involved process, but given the amount of investigative time that had gone into this case, now that's a drop in the bucket in terms of building back that tree. What they ended up with was narrowing the case to about five male individuals. There was one in particular in that list that really stood out. He was a former police officer. There were some other indicators in terms of why they wanted to take a close look at him. They took a covert sample and they could not exclude Joseph James D'Angelo from the original crime scene evidence. He has since gone to trial. He entered a guilty plea on 13 counts of murder and 13 counts of kidnapping. Some of the other crimes had reached their statute of limitations. But that crime was solved. It attracted a lot of attention, and it gave some closure to the individuals, the relatives and the victims of the Golden State killer.

Nathan: So that case, those two cases are very early cases. This is just a chart showing the use of this technique by law enforcement since that time. This is courtesy of a project in Mendeley that is tracking all of the cases they can identify where the technique's used. So you can see a steady stream of cases over the last three or four years. The technique is increasing in terms of interest from law enforcement, but it relies on that support from the public in terms of making their data available for comparison. That process, of course, raises some privacy and ethical concerns, which we will come to shortly. So with that, I'll hand back to you, Marie.

Marie: Thank you, Nathan. So many of our attendees today may have taken a genetic ancestry test with a direct to consumer genetic testing company. Does that mean that the DNA is being used for forensic genetic genealogy, whether they like it or not?

Nathan: Oh, thank you, Marie. It's important to note there are only two platforms, and they're some of the smaller platforms that actually permit law enforcement use of that data. And that's the site called GEDmatch and Family Tree DNA. And both of these sites now have an opt in, opt out arrangement for criminal use of samples. So users do have the option of allowing their DNA to be used by law enforcement in this way or opting out of that process. GEDmatch does allow some wider use for human remains and information only. For those who opted in, though, the law enforcement do not have access to the raw DNA. They only have access to information about how much DNA they share with it, with any particular sample that's uploaded. So if they share any DNA at all, that will appear in an extensive list of potential matches. But the technique also does rely on public records. So it's important to note that individuals who publish their family trees and sites like Ancestry could form a critical step in the investigative process. That information might help construct trees just like it did in the Buckskin Girl case. So those trees are being used without explicit consent. They're effectively being treated as published information. Maintaining that public trust and confidence is critical to this capability because if people are concerned about this data being misused in any way, they will simply remove it from these databases. They'll hide their trees. They'll opt out of law enforcement matching. Some of these issues extend beyond law enforcement use. And certainly people who are providing any sample to a commercial database need to be aware that other individuals, aside from law enforcement, can match with their particular genetic data. And there have certainly been individuals who have had surprises come up in their in their family trees after uploading to one of these platforms. Back to you, Marie.

Marie: Thank you for that. Dennis, for those companies that do allow law enforcement use of their holdings, how extensive is the reach of forensic genetic genealogy? Can we really find a genetic relative of any donor of the crime scene DNA sample?

Dennis: Yeah. Thanks, Marie. So this scenario has been modelled and what we see on the screen now is, on the left hand side here, you can see a plot that shows the probability of finding a match on a typical forensic genetic genealogy database based on the degree of genetic relationship. So over on the right here, we have our first cousins once removed. And, you know, the probability is very low of finding a match. And so it would be even lower for a sibling or a parent. But as we move out towards more distant relatives, like second cousins, third cousins, fourth cousins, etc., then that probability of finding a match increases. So it's very, very likely that almost 100% likely, that you'll find a fourth cousin on these databases. But of course, as Nathan has explained, that's a very distant relative and not always useful because the family trees that need to be constructed are quite extensive. So over on the right here, you can see some modelling that shows how big these databases need to be in order to find a genetic relative. So the green line represents the probability of finding a first cousin. The blue line is the probability of finding a second cousin. The brown line is the probability of finding a third cousin. And the black line is probability of finding a fourth cousin. And you can see that you only need about 2% of the population, that's any population, to be present in these databases before it becomes very close to 100% probable that you will locate a third or a fourth cousin. And certainly third cousins are very useful. And that provides quite a likelihood that you'll be able to identify genetic relatives. So 2% seems to be a ballpark figure in order to find anybody from their genetic relatives in these databases. So are we there yet? Well, let's look at the holdings of these databases in the US because that's where these companies are and where the majority of the subscribers are. And you can see that the biggest direct to consumer service provider is Ancestry DNA with over 20 million profiles in its database followed by 23andMe with over 12 million. MyHeritage with five and a half million, FamilyTreeDNA with almost 2 million and Living DNA with a little bit less. So in total, we've got over 40 million profiles available in these databases. Now, 40 million is about 10% of the US population, so well over that 2%. So the potential is certainly there for identification of just about any unknown profile. But of course, we have to remember that not all of these profiles are available or accessible by law enforcement. And as Nathan has explained, there's only a couple of providers that provide law enforcement access at the moment. One of them is FamilyTreeDNA. The other is a company called GEDmatch. And so when we look at the holdings of these two providers, with each of them having well over a million subscribers, again, you can see that we've got over 3 million profiles that are in these two providers that are potentially accessible by law enforcement. Now, 3 million is about 1% of the US population. So you can see that we are certainly getting near that ballpark figure. But the other thing to remember is that not all of the of these profiles are available. So for FamilyTreeDNA and GEDmatch, subscribers need to opt in, so they have to choose to make their DNA available for law enforcement use for violent crimes at least. So it's going to be an even smaller fraction of this 3 million that have made or opted in or chosen to make their DNA available for law enforcement use. So not at 2%, but you can see the trajectory is increasing. There are increasing numbers of subscribers to these service providers. In fact, I heard the other day that forensic genetic genealogy, anyway, is the second most popular hobby in the world after gardening. So certainly there's potential to reach that 2%. Thanks, Marie.

Marie: Thanks, Dennis. Interesting. Are there any other uses of forensic genetic genealogy beyond catching killers?

Dennis: Yes, there are. And as Nathan has explained there are some ethical concerns about the use of these databases, and surveys have demonstrated that there is some polarisation in the community about those who think that DNA should be used for law enforcement to find perpetrators of violent crimes, and those that think that they shouldn't be or that there should be some caution. Now, one use of these databases seems to have fairly high levels of public acceptance, is to use them for missing persons investigation. So not criminal matters, but but coronial matters. Looking at missing persons and unidentified human remains. Now, there are a lot of programs around the world that are involved, that are dedicated towards this endeavour, including in Australia, the Australian Federal Police, National DNA program for unidentified and Missing Persons. And certainly, forensic genetic genealogy is one of many techniques that this national program are using to help identify human remains and bring closure to families who might have missing persons. There's a couple of caveats, though, when we're talking about unidentified human remains. And I want to begin by talking about how these DNA profiles are uploaded to the genetic service providers, how they're produced, and they're produced using a technology called microarrays. And on the left here, you can see this is what a microarray looks like. It's basically about the size of a large microscope slide, and it consists of about a million wells, from half a million to a million tiny little wells etched into the surface of this plate. And each well is responsible for returning the DNA profile at a very small location on the genome. So, in total, we're surveying about half a million to a million locations on the human genome. So, we're not sequencing the whole genome. We're just sampling from it, about half a million to a million locations on the genome. But that's enough to create these identity by descent segments that we're interested in. And these microarrays can then be read on an instrument like the one at the right here of this this picture to create these DNA profiles. Now, as I've mentioned before, you need about 10mLs of saliva to provide enough DNA to do that. And it has to be good quality DNA as well. But the problem with missing persons is that we're often dealing with skeletonised human remains and we're trying to retrieve DNA from a bone fragment like the one shown here. And unfortunately, DNA is very difficult to, in many cases, it's very difficult to retrieve from bones because of the matrix that it's encased in. The bone is made up of calcium, and calcium can interfere with downstream analyses. And not only that, but the DNA is often degraded, especially if the human remains have been exposed to the environment for many years or buried in a clandestine grave. But it can be very difficult to obtain high quality DNA. And so the microarrays don't often work. So there's a lot of research being looked at, trying to create DNA profiles that are sufficient for uploading to these genetic genealogy databases using other technologies, perhaps whole genome sequencing or a technique called targeted amplicon sequencing. So, yes, missing persons investigations and unidentified human remains is certainly another application for forensic genetic genealogy. Back to you, Marie.

Marie: Thank you Dennis. So, Nathan, is everyone on board with forensic genetic genealogy? Why might someone be wary of signing up for law enforcement use of the DNA?

Nathan: Thank you, Marie. And certainly there are a number of considerations in uploading DNA and providing consent for law enforcement matching. We need to distinguish between those broader issues around simply providing your DNA to a company. And I've already touched on the unexpected surprises that individuals have come across in their own family trees. You're also trusting a company in terms of data security and privacy to safeguard that genetic data. But that applies irrespective of whether you allow any law enforcement matching. When you come to law enforcement matching, there's obviously a lot of good that can come of that in terms of solving cases. The data we're seeing from GEDmatch is that over 70% of new subscribers are opting in to allow their data to be used in this way, but there are some risks and concerns. So forensic genealogy really is a family donation. And law enforcement investigations could implicate a family member in a crime. It could also be used to try to identify a distant relative who is unidentified in terms of being a missing person. Now that in terms of the criminal aspect, could be a very distant relative, but it could also be someone closer to you. So most of the cases we've seen involve second and third cousins of any of the matches. Not too much closer. But the Buckskin Girl is the one exception where those individuals who uploaded would have been aware of that individual in their tree. It could even apply to individuals who are not yet born yet. So those individuals, once they're on once your DNA is on that database, they are within reach of this technique. The technique also doesn't know any geographic borders. So once you opt in, you've opted in for use worldwide. And we know this technique is more common in the United States at the moment. And that brings into play issues around the death penalty and cases where individuals could be subject to the death penalty if convicted. There are many, many individuals who are happy to assist law enforcement, but would be concerned if that assistance resulted in a capital investigation. There's also a lot of controversy around Baby Doe cases. Now, these are on the face of it, cases where we're looking to identify an infant or a baby. Of course, that identification almost always then results in a close family member being a suspect in their death. So that is one that cuts across between human remains, identification and the criminal side and very hard to differentiate between the two. So my research, a work in this area really has been about developing policies and strategies that can ensure that this technique will be used appropriately and will maintain that public trust and confidence. Part of this is ensuring that there is awareness of what this technique can and can't do, and that applies as much to investigators who may be receiving intelligence leads from forensic genetic genealogy. There's always more work to do after that initial lead is generated. It's never enough to go and lock someone up. It's never enough to go and knock on a door and give a family the news that their loved one's been found. This is one piece of a puzzle. It can greatly assist law enforcement, but then needs to be followed up by traditional forensic and investigative work to ensure that's a sound conclusion. It's a technique that some have said will end the serial killer, that it will stop these recidivist offenders because we'll be able to cut them off after a small number of offences rather than allowing them to continue to offend for decades. We'll see whether that holds true. But it does pose, certainly it is of great interest to law enforcement around the world, but needs to be implemented in a very careful way to ensure we keep the public on board. Thanks, Marie.

Marie: Thank you so much, Nathan and Dennis as well. I'll just share my screen again and we're going to go for the Q&A session. So we had quite a few questions already on the chat and also at registration, so we'll try to go through them during the time remaining. So I had a look at the question and there is one in particular that attracted a lot of attention, and I think I'm going to give it to Dennis. What type of jobs are available in forensic science area after graduating?

Dennis: So are we talking about jobs that could involve forensic genetic genealogy or just any jobs in forensic science?

Marie: I'd say it's more general, but you can probably focus on forensic genetic genealogy, maybe.

Dennis: So, I mean, obviously, forensic science is the use of science to help solve legal questions and questions of interest to the courts. And so the destination of choice for most of our graduates is with either a police jurisdiction or in in states where forensics is conducted outside of the police, a forensic laboratory. So, you know, in Australia, there are some states where forensics is conducted within the police, like the Australian Federal Police and Victoria Police. And there are other states like New South Wales where forensics is conducted inside another state department, Department of Health, for example. So that's obviously a prime employer of forensic graduates. But forensic science is really, it's a science. And so our graduates are also very qualified to work in a number of other settings in any laboratory based setting, for example, but also professions that require critical thinking and analysis and the gathering of intelligence to try and answer other questions, not necessarily legal ones. So I hope I've addressed that question.

Marie: I think yeah, I think it's fine. If they have more questions, you can always ask at a later stage. The next one, it's also related to degree. It's probably from a student. If I'm majoring in CSI, do I have to work as a police officer before working in crime scenes, or is there a different route that can be taken? So maybe for Nathan this time.

Nathan: In most jurisdictions, no, there's a separate intake of forensic members into crime scene areas. It would depend on the state and territory. But by and large, there is a direct entry program out of forensic university programs into that field.

Marie: Great. Thank you. Then I've got a question from Claude. Ethical concerns and public confidence are often stated as critical consideration before this approach can be applied. Not saying it's not important, however, conceptually, how is it different from investigators searching out of reach through databases using an incomplete number plate followed by traditional investigating techniques? So maybe, Nathan, again, I'd say from the law enforcement perspective.

Nathan: And there are certain elements of this approach involving the genetics that make it more privacy intrusive. So we need to factor in the DNA specific aspects when we consider the privacy implications that we are looking at family relationships, these there are sensitivities around families, family groups, cultural sensitivities that we need to work through. So it's not quite as simple as some of those other investigative techniques. And yes, there are many, some are not very intrusive, like searching a number plate. Others may be more intrusive, like pinging mobile phone towers or carrying out other forms of more invasive surveillance. But we need to ensure that with all these capabilities that we've considered the privacy implications, that we've got the frameworks and governance in place to make sure that they're being applied in the most appropriate cases, and therefore, that we can maintain that public confidence in any police technology that we might use.

Dennis: I mean, if I can just add to that one, Marie. I mean, I think the question from Claude highlights an important consideration, that there are similarities with other investigative techniques. So, for example, one of the first ports of call in many investigations these days is social media. And a lot of familial relationships can be found in social media, you know, including Facebook, photos of marriages and where everyone's tagged in the photograph. And so a lot of these familial relationships can be also made available from other sources. But as Nathan points out, there are some aspects of DNA that require extra attention.

Marie: Thanks for that. The next one is how far back in family history does the DNA Ancestry results show? So I think, Dennis, you mentioned that as well in your talk. So maybe a question.

Dennis: Yeah. So I've only just been made aware recently thanks to Nathan about a couple of cases of forensic genetic genealogy that have been applied to historical crimes. One about a hundred years old, post-mortem interval, and one about 150 years old. So certainly that range is possible. You have to remember that the farther you go back in time, the more removed the deceased person, or the deceased person in many cases, is from the most recent common ancestor. So the family trees involved are going to be very extensive the further you go back in time. So that perhaps creates a limit as to how far you can go back using forensic genetic genealogy, remembering that, you know, that ten centimorgans is probably the limit of detection that we can apply to forensic genetic genealogy. So once you get to three or four times removed for a, for a relative, it becomes quite difficult to construct that family tree. Most of the crimes that have been solved, the cold cases using forensic genetic genealogy are really within a few decades. But certainly it is possible to go back further.

Nathan: I'll just add to that as well, if I may. In terms of the limits of this technique, certainly around ten centimorgans. Those types of matches are still actionable. They certainly can assist in reinforcing your existing hypothesis because effectively you're triangulating matches back in time. So it's not uncommon with this technique to be building trees that do go back to the 1700s, sometimes even the 1600s, and linking together quite a number of matches to try and work down to the present day. You don't know until you upload the sample whether you're looking at a case like the Buckskin Girl with a potential first cousin once removed, or whether you're way out in in terms of potential matches. That's one area we're looking at in Australia is just how effective will this technique be when we upload any samples? Are we going to get close enough matches on a regular basis or are the databases just not quite, don't quite yet contain enough Australian profiles that we're going to start to get closer matches, which makes it more efficient?

Marie: Thank you for that. And another question for Nathan this time, could the providers be ordered by court to make the records available regardless of opt in options or policy of providers?

Nathan: It is possible for a court to order the disclosure of any information subject to some kind of warrant or subpoena. The larger providers like AncestryDNA have had some requests through legal process and they've resisted all of them. I would imagine a company like Ancestry and 23andMe would be taking their own action to protect their client information if there was some kind of warrant or subpoena issued requiring them to hand over or search their database. So apart from those very specific instances where it's been tried and then quickly overturned, I'm not aware of any case where that has resulted in information being released. It may well make its way right up through the courts if anyone ever tried to push that that line too hard.

Marie: Thank you, Nathan. Dennis, you mentioned in the beginning of the presentation that DNA can be used to determine geographical backgrounds, ethnicity of an individual. Is this side of the technology used in law enforcement, or is it mainly to try and find genetic links with DNA databases?

Dennis: So if I understand the question correctly, yes, it is used or can be used by law enforcement. So the same algorithms that are used to find genetic ancestry can be used to identify genetic relatives. But perhaps the question is asking, is the genetic ancestry of an individual useful for a forensic investigation? And the answer is yes, because you can use that information to narrow a pool of potential suspects or candidates just as much as you can by finding genetic relatives. So if you are able to determine that the donor of a crime scene that you're interested in, a person of interest has Irish ancestry like I do, then you can eliminate from that investigation any suspects that you might have that have other what we call biogeographical ancestries. So someone who has essentially an East Asian by geographical ancestry or an African by geographical ancestry, and then concentrate your efforts onto those individuals whose ancestry is predicted by these algorithms.

Marie: Thank you. And I think we have time for one more question. I'm sorry if we didn't get through all the questions, but there was quite a few. So I've just got to ask Nathan, would this technology be used for events like 9/11 with bone fragments?

Nathan: It certainly can be used in disaster identification type scenarios. In fact, one example is the campfire wildfires in California a few years ago. Most of the identifications were made by close family members providing DNA, which was then compared to the deceased individuals. But there were a small number where there wasn't a close enough relative to compare to. And in those cases, this exact technique was used. Building back their family trees and identifying and building down to relatives who provided samples in those databases. So they provided that extended reach that was needed to inform the coroner, they already had other circumstantial evidence as to where the deceased was found and so forth that supported identification. But this this technique was then used to provide that additional evidence to a coroner that led to those identifications.

Marie: Thanks a lot. And again, I'm really sorry if we didn't manage to get through all the questions, but I want to thank again the two speakers today for this really interesting and really lively talk. I learned a lot myself, so we will send a copy of today's talk along with the Q&A session to everyone who registered. And it will also be on our website. So you've got the website here on this slide. So if you wanted to have a look at it at a later stage, you are welcome to do so. And just before I leave you today, we've heard some really innovative approach to forensic science. And amongst other things, it has shown how forensic science can capitalise on digital transformation and technology to better integrate with investigations. These topics will be further debated with world experts at the forthcoming IAFS conference in 2023 in Sydney. The meeting is in good hands because Dennis is one of the discipline convenors for that conference and so make sure you stay tuned if you are really interested in that field. And I think I'm going to close the session today and have a great day, everyone, and thanks a lot for coming along and thanks again to the speakers.