Applying Ancient DNA Techniques to Forensic Samples – ISHI News

May 13 2021

Applying Ancient DNA Techniques to Forensic Samples


In this interview, Kelly Harkins Kincaid, co-founder of Astrea Forensics shares how her background in paleogenomics makes her uniquely qualified to tackle degraded DNA found in casework. She also discusses why it was originally thought that there was no nuclear DNA present in hair as well as what makes hair such a great sample type for forensics cases. We also examine how Wooly Mammoth samples from an ancient DNA research lab have opened new doors for solving cold cases.

To learn more about this work, watch our interview with Ed Green.






Laura: Hello and welcome to the annual video series from the International Symposium on Human Identification. Today, we have Kelly Harkins Kincaid joining us. Kelly, why don’t you tell us a little about yourself?


Kelly: Great, thanks for having me. It’s great to be here (virtually). I am CEO of a new startup called Astrea Forensics. We’re about a year old, and what we do is provide a service for law enforcement and related law enforcement agencies where we provide whole genome sequencing data from degraded samples for the purposes of investigative genetic genealogy. So, this is born out of our work collectively; myself and the founder out of the ancient DNA research world. We found a really great foothold here and thought it would be a useful service to provide, and so far, so good, so I’m excited to be able to talk a little bit about it.


Laura: We are so happy to have you. I know you’re doing some really interesting things. First, tell us a little more about how Astrea got started. What was the genesis?


Kelly: Yeah, so I’ll go back just about a few years for myself and will then talk a little bit about Ed Green, just very briefly. So, my background is in paleogenomics and my research is focused on getting miniscule pathogen DNA; reconstructing pathogen DNA out of archaeological human remains. So, my background is very much what would be called paleogenomics or archaeological genetics. So, I’m very used to dealing with samples that are at baseline degraded and attempting to get something even smaller out of them. Most of my research questions were very anthropological in nature; understanding disease transmission and origin in prehistoric populations. But, when I began a post-doc at the University of California Santa Cruz in the USC paleogenomics lab, I got really drawn in by methods development, so it brought me onto a different kind of path. Ed and I have been working on some methods that we thought would be really useful on samples other than ancient DNA, and we spun up a company called Claret Bioscience in 2017, so I established that lab and facility and had been leading a team of about 8 or 9 folks there when our technology started to be used for a different purpose, which is forensics. So, the technology we offer is this great way of getting really degraded DNA out of samples and getting it sequenced and folks started to reach out to Ed Green (who is still faculty at the university even though he’s founder of these companies) and say, “Hey, I have these really terrible samples” or “I’m working with hair and I hear there are ancient DNA techniques that work with hair. Can you help me with these cases?” At his lab on campus, all of a sudden, he had all this interest from a totally different part of the company. Meanwhile, down the hill, I was already working in a commercial lab with Ed, and given my ancient DNA background, it just was a real nice melding of our experiences and the kind of infrastructure that we already had established, and my ability to set up a lab really quickly and get it running. So, that’s how it started. That’s a long answer to a question.


Laura: No, that’s amazing! You’ve been a little bit busy!


Kelly: And growing, so it’s great.


Laura: Let’s talk more about the degraded samples. How do you work with them? What are some of the benefits, challenges…


Kelly: One of the sample types that we excel at and is one of our more unique capabilities is working with rootless hair strands. Of course, we can work with bone and teeth as well. We tend to apply ancient DNA techniques to them. So, techniques that are designed to get the shortest possible DNA fragments out of a sample. The benefits to using something like hair versus bone and teeth is that human remains often take organisms from the environment in which they’re buried, because they’re kind of a porous substrate, whereas hair tends to not be as permeable to micro-organism contamination. When you get a bone or a tooth and you get all the DNA pieces out of it that you can, and you sequence them, sometimes 99% of those DNA fragments aren’t from the human themselves. This is very different in hair. There are cases where hair is just extraordinary preserved. We’ve tested some things in an ancient DNA world, like Wooly Mammoth or Wooly Rhino. These are extinct mammals that have been extinct for 50,000 years, and the DNA degradation in the hair strands is the same kind of degradation as the hair that is on your head right now. So, applying the ancient techniques and performing whole genome sequencing is kind of the paleogenomic pathway that we’ve applied to these other samples.


Laura: That is amazing what you’re doing with hair, and that was a great illustration of what it is for anybody who’s watching this and has worked with that before. We’ve certainly had a lot of people from the ISHI conference who are an expert in some of these things, but then we have a lot of people who are just interested in the field and want to learn more about it. Maybe we can talk a little about mixture deconvolution too. I think you’re working with that?


Kelly: Yeah, there’s one other point that I’d like to make about hair before I move on, because it occurs to me (like you said) that there’s a lot of folks who have been in the field of forensics for a long time and have not formally worked with hair, so the question is why not? The thing about hair is that the hair on your head right now, as I just said, gets degraded almost immediately. Once it’s degraded, it maintains this level of degradation for a long time, which is one of the benefits of working with this type of sample. But, it degrades so short that the DNA pieces are too short for STR profiling or any PCR-based approach, because the amplicons those methods are trying to target are just longer than there are pieces of DNA in the hair strand. So, for a long time, it was thought that there wasn’t a lot of nuclear DNA present. It’s not that it isn’t present, but it’s too short to be observed by those types of methodologies. So, for all intents and purposes, it wasn’t there. It was observable, and it wasn’t useful to anyone. So, once you have a method where getting a 30-base pair DNA fragment, and getting millions of them over the whole length of the genome; once you have a method that can do that, then you can sort of reconstruct a lot of the genomic information. So, I just wanted to point that out, because a lot of folks are wondering how it is that we’re getting nuclear DNA from hair, and it’s certainly in there, you just have to use a different method to get it out. We do use published methods for the DNA extractions and the library preparation, so those are available online.


Laura: Thank you. That is amazing how far we’ve come and how many things we can put together now to do different techniques. A puzzle is a perfect way to say it.


Kelly: Let’s talk mixture deconvolutions. So, one problem that we’re addressing and think is a very difficult and interesting problem is when you have a sample that has multiple contributors. The beauty of hair is that it’s a single biological unit. We only ever process one hair at a time and it could only have come from one individual. So, that’s really nice. There are many cases where you’re going to get a sample that had multiple contributors. So, right now, we’re working on a computational method. Right now, we’re only working with two contributors, but the idea is that we’re going to be able to use population genetics as a framework for separating out those two individuals. But, not just separating them, but actually providing an entire genotype file for the unknown component. So, assuming that, in a situation where one component is a known victim and another component is an unknown person, the idea is to generate the entire genotype file so that file gets treated like any other genealogy-based output. We think that once we’re through the R&D phase of that, it will be really exciting and valuable.


Laura: That’s incredible, yes! How about ancient DNA? You touched on it, but let’s talk more about the work you’re doing there.


Kelly: Having a relationship with the university and the paleogenomics lab is useful and is churning out great stuff all the time. We have a really great relationship where we can go back and forth on methods development with them and have the benefit of being able to access some of the samples that they maybe have some excess of where we can play around with some methods in development (for example, the ancient mammoth and the ancient rhino). So these help us to really make sure that our methods are robust and we’ve learned a lot, of course, about how DNA degrades by just understanding samples from all kinds of time periods. From all kinds of organisms. I think the oldest humans that I’ve worked with are 10,000 years old all the way up through the modern time periods. But when you’re using extinct animals, you get to go even further back, so it allows you to understand DNA degradation rates and what happens with something called post-mortem damage. What features get changed when damage occurs in the molecules? This actually happens kind of immediately. In your body, when DNA gets damaged, you have biological machinery that goes in and fixes it. The moment you cease to live, all that biological machinery is done, so that’s how DNA starts to degrade from the moment you die until however long. So, we’ve been able to use ancient DNA as a learning tool. Of course, the methodologies that come out of ancient DNA are so useful for forensic samples. We use the methods that are meant to get DNA from (in some cases) hundreds of thousands of years old. So, when I moved from ancient DNA to this forensic context, these samples were no problem! There are some exceptions. There are always samples that give you trouble, but it’s a nice way to apply a research based strategy to something that matters in the real world. I felt really at home since starting this company.


Laura: I think that’s amazing and what a progression. Taking those techniques from ancient DNA. I know talking with Ed, we spoke to him three years ago when he was working on the neanderthal genome and projects there, so the history that you both have in this company is fantastic. What’s next on your radar?


Kelly: That’s a good question, because the other side of the company is developing the same techniques we apply in a forensics context to clinical samples. There are a lot of clinical samples that kind of mimic and have some of the same characteristics as an ancient or degraded sample. So, we’re trying to find different applications for these same techniques. The ancient DNA world has all of these great methods that turn out to be useful across many different sample types. So, we’re working on some clinical stuff. We’re even working on some of the Sars CO-2 enrichment. So, that goes back to trying to find the needle in the haystack of a little virus genome in a mixture of things that are human. In the ancient forensics lab, we’re hoping to get the word out there and we realize that hair is often collected in different cases, even though it may not be the first sample type that a crime lab goes after, so we’re hoping to make this service as broadly useful as possible to as many folks as possible. We think it’s a great use of the technology and we’ll work towards accreditation. Right now, we’re a research lab and are very young, so we have terms and conditions that try to follow very closely with the DOJ guidelines for the use of this data for forensic genealogy. Hopefully we can kind of grow past that as we grow with the field as the field grows and be there on the forefront.


Laura: Well, I feel like we’ll definitely want an update next year and will keep following this. I think it’s going to be fantastic. Kelly, thank you so much for working through our virtual interviews this year.