Jul 06 2022

FGG Webinar Follow-Up: Rachel Oefelein Answers Your Questions

In Part 2 of our investigative forensic genetic genealogy series, held on May 4th, Rachel Oefelein, Director of Research and Innovation, Quality Assurance Manager, and a Senior DNA Analyst, DNA Labs International, presented a step-by-step guide from DNA extraction through prosecution of Forensic Genetic Genealogy (FGG) cases. Ms. Oefelin also addressed how to generate a profile suitable for genealogy upload, the bridge between generating genealogy data and genealogy research, confirming investigative leads with STR comparison, and expert witness testimony as it pertains to cases that utilized FGG.

Like leaves falling from the metaphorical genealogy tree, sadly not all questions can be answered in the allotted time during a webinar. This blog will highlight some great questions that have not yet been addressed. If you missed any of our three-part webinar series, you can watch them on demand.

Do you anticipate standardization of the field of forensic genetic genealogy?

Absolutely! Forensic scientists like rules and to provide assurances to our stakeholders (criminal justice community, general public, policy makers). We need to ensure that this invaluable tool is being used appropriately. History has shown that when there are not effective checks and balances in place that is where the risk for miscarriages of justice exists. Policies take time to enact but the Department of Justice has already provided a starting point with the UNITED STATES DEPARTMENT OF JUSTICE INTERIM POLICY FORENSIC GENETIC GENEALOGICAL DNA ANALYSIS AND SEARCHING released in 2019. I suspect we will see additional standardization beyond that in the coming years.

How do you think FGG will change in the coming years?

In addition to increased standardization, I think we will see huge improvements to the technology that will be more forensic-centric, opposed to the diagnostic/research/commercial-based assays we have been used to. The biggest change I am looking forward to is the increase in the database size of the public databases we are using, FTDNA and GEDmatch. Increase in database size will lead to faster processing of FGG cases and reopening cases that previously had limited match lists.

What’s one FGG case you wish you could work?

It’s hard to pick just one! Natalee Holloway’s case speaks to me just because there has been so much press from her family, and I think about putting myself in their shoes if it was my loved one. It would be nice to provide even a little bit of closure.

Are the SNPs used in the Kintelligence kit compatible for purposes of comparison with ancestry tests performed by other platforms?

Yes! Kintelligence SNP profiles are uploaded to a public database the same as any other kit. They are classified by case type which limits the search results, much like any other law enforcement profile. Match lists that are returned contain kits from all other types of uploads that were opted-in for law enforcement. Or, in the case of GEDmatch, match lists are not restricted in John Doe and Jane Doe cases.

Did you say you can test fixed samples? Would that include formalin samples and samples made into either wax blocks or glass slide?

Yes, these are obviously challenged samples for a battery of reasons; however, we have an effective protocol in place that has had great results in these cases. I find often these cases are rejected outright but the process of fixing the samples is directly correlated to the success of the sample. For example, if the fixing procedure was less than adequate, we have a better chance of results. The trouble with wax block samples is that they are often prepped in mass and then cut so you need to be wary of mixtures with adjacent samples. For these, it’s imperative you utilize a substrate control as well.

Can files generated by Kintelligence be uploaded to FTDNA?

As of now, no; however, this is an issue that I expect to be resolved imminently.

As part of your validation, did you genotype the same sample with a traditional genetic genealogy tech (generating >600000 SNPs) and with the Kintelligence kit? How did the GEDmatch/FTDNA match lists compare?

Yes, we compared to WGS and SNP Array data. Most differences are the loss of a sister SNP, and this was seen sporadically throughout all three systems. There are also instances of null SNPs or no result SNPs in one system whereas in another system it’s called. There were not significant differences in close matches, where you see differences is in degrees of relatedness further out, like 3^rd cousins and beyond. However, the recent algorithm updates in GEDmatch have resolved many of those.

How successful are your cases if you are only able to identify up to 3rd cousins?

Depends on the case! We have cases that are quite simple and cases that require considerably more research. However, I would say that is true with any of the systems. We do not solely work Kintelligence profiles, we also work with SNP array and WGS data. I would argue that 4^th cousins and beyond require considerable research regardless of the system. With thousands of cold cases and unidentified human remains cases every year, I believe there is value in triaging. We can spend limited government resources on a case with only distant cousins and possibly find success months or years down the road, or we can continue to work and upload other cases with closer matches and pause on the remaining cases while the databases grow. It is two different schools of thought and I think it is a conversation that needs to be had jurisdictionally.

It sounds like the NGS, or lab portion of the lab was accredited. What did the auditors say about the genealogy needing to be accredited?

Our protocols under the scope of accreditation cover everything through upload, and the genealogy research is not included in what falls under the scope of our accreditation. Typically, auditors are to advise when you are not in compliance with a standard, so if there is no existing standard that we are auditing against that would be without comment. At DNA Labs International, we are taking a proactive stance and voluntarily adhering to what may be considered as standards moving forward, for example, some of the FBI QAS analyst standards like literature review and continuing education. Additionally, our genealogy case files undergo reviews as well.

Does Kintelligence work with rootless hair samples where the highly degraded fragments are only ~40-50 long?

I would say if the fragments were only 40-50 base pairs long that I doubt it would be successful. With that said, I haven’t tried it that low. We have rootless hairs that give partial and full autosomal STR profiles, in those cases, if there is enough extract remaining, I would expect a better chance of success.

Do you have any recommendations for how a forensic scientist can train on Kintelligence on their own so they can effectively present this technology as an option to their laboratory? How would they get hands on experience with the data interpretation?

There are a lot of great webinars available online. I would also recommend talking to your local sales representative for additional resources.

There should never be a reason for the genealogist to testify.

There have been genealogists that have already received subpoenas for court. I would encourage everyone to review fruit of the poisonous tree doctrine (the Derivative Evidence Doctrine) and the fourth and sixth amendments. If the admissibility of the genealogy is called in to question (and admissibility hearings do require testimony), any evidence obtained because of that genealogy research or evidence as a result of warrants obtained because of the genealogy research may be called in to question. Bottom line, if you work on criminal casework there is a chance you will have to testify. I am not an attorney though, so I would always recommend consulting with your local district attorney, state attorney, etc.

What is the likelihood that more common (i.e., less violent) crimes will be addressed with FGG?

I would say we are far away from that if ever. Genealogy would have to cost considerably less, and you would have to get stakeholders, database owners, and the general public on board, which is not a small list!

What are the chances that CODIS will open to matches beyond direct familial matches?

I would say unlikely, even existing familial searching can generate large match lists and become extraordinarily expensive. Also, we would need more locations likely and many of the profile’s pre-date the expansion of the core loci in 2016. Furthermore, most states do not even allow the current iteration of familial searching.

How does law enforcement afford the cost of these services outside of CODIS?

Many ways! Crowdfunding, grants, donations, funding from smaller municipalities, etc.

Would Kintelligence be appropriate for an unidentified deceased infant case where all we had left from evidence was a blood smear on a card from the state lab? We pulled a sample, but it was pretty degraded. We got results, but our genealogist is struggling because the closest we can get is 3rd cousin.

It would depend on the previous results. If you had a great result the first time, Kintelligence will not increase your match list. If your previous profile was limited, then it may be beneficial.

Is not having access to FTDNA is major limitation and do you know if this can be addressed?

My understanding is this will be addressed shortly!

Did you say that your team solves a case in 4 hours?

Not every case but yes, some cases! In fairness, that does not include the amount of time it takes for the profile to upload but rather actual hands-on time.

What kind of SNPs does the FGG use? mtDNA SNP, or Y SNP or autosomal SNP?

Kintelligence SNPs are classified as Ancestry, Identity, Kinship, Phenotype, X-SNPs, and Y-SNPs.