Scotland Yard Leading Pilot Program in IGG

[vivazapata]

au.finance.yahoo.com/news/killer-family-tree-dna-sites-094938765.html



Police are exploring a DNA technique which is credited with cracking scores of unsolved murders in the US, the Standard can reveal.

Scotland Yard said it is leading a pilot on behalf of national policing into investigative genetic genealogy, which compares the DNA of an unknown suspect or unidentified victim with samples uploaded by the public onto a genetic genealogy database for likely relatives. It allows genealogists to place relatives into clusters of family trees and work backwards to identify them.

It could be more effective than the gene-tracing techniques used with the national DNA database as it can trace a potential suspect from more distant relatives.

It could lead to genealogy sites being asked to open up their databases to investigating officers. The technique could uncover a person’s family tree if someone as distant as a third cousin uploaded their DNA onto a database. At present, the UK’s database used by police is only searched for close possible relatives such as parents, children or siblings to match a suspect’s DNA.

The genetic genealogy database is credited with solving the Golden State serial killer case in California in 2018 among other high-profile examples in the US.

Joseph DeAngelo was convicted of at least 13 murders and admitted dozens of rapes in the Golden State case after about 20 people were identified as third or fourth cousins to the perpetrator through the technique.

Deputy chief constable Ben Snuggs said: “Matching DNA to relatives, investigative genetic genealogy can lead to the identification of a perpetrator.”

He said that a working group had been established to “explore the practical and ethical implications of genetic genealogy and assess the potential of this technique for policing”. However, he said no decisions had yet been taken.

Debbie Kennett, an honorary research associate at UCL, said there were fewer cases where it would be used in the UK because of the numbers already on the national DNA database and fewer unsolved cases.

But she added: “Even if it was used on those small number of cases, they could potentially be solved.”

The project comes after a Government paper in 2020, which warned of potential privacy concerns.

The Home Office-commissioned report advised that “unregulated” practice of genetic genealogy database should only be used if necessary and when other methods have been exhausted, and once approved “by the appropriate body”.

The most popular databases, like Ancestry, do not allow police access to solve crime. Others like GEDmatch allow user opt-ins for serious crimes.

[sammyt]

I thought they had already used this technique in UK cases, such as the Geoffrey Godfrey rape case.

He was caught after a 16 month long familial search, conducted by police, genealogists and civilian detectives.

[vivazapata]

I am not 100% familiar with the DNA situation in the UK. In the US you have 3 types:

CODIS the system that states upload their DNA into. It is a national system and uses the LE 20 loci/point system to ID. This was how Christopher Ewing the Colorado Hammer Killer was ID'd. Different states have their own criteria as to what crimes qualify for inclusion in their database and thus what ultimately goes into CODIS.

Familial searches are done(in the states that allow it) by comparing the LE 20 point samples and generating leads that match some but not all the points within their respective database. In California this method was made possible by a bill basically funded by Keith Harrington's brother Bruce into a bid to ID the EAR/ONS. JJD wasnt in the database but the method did ID the "Grim Sleeper",serial killer Lonnie Franklin Jr.

This seems to be similar to the UK system from what I read on the wiki. Some of the collection seems to be the same but from articles I have read they seem to do DNA sweeps gathering (voluntarily?) groups of samples, maybe this is done more/only in areas with a more static population.

IGG uses a full autosomal profile which contains many more comparison points like hundreds of thousands or millions. In essence you are casting a much wider net and then it allows you to cross reference (at least in GEDMatch) to see the degree of separation.

Where the government databases work by exclusion IGG works by inclusion. I do not know what database they would use whether that would be GED match or if there is a equivalent for UK/Europe. Even if not my guess is GED match could be a good resource as I have it on good authority many of the people uploading their data may have come from Britain-lol.

According to the article I read on the case mentioned ,however it was collected, a relatives DNA was a partial match. They then created a family tree with 4,000! people to find the suspect.

In this case they in essence got lucky. One of the first stabs at IGG was in the case of Angie Dodge. One of the detectives ran the DNA against a database that existed at the time called the Sorensen database using a form of DNA known as a Y-STR sample (male and can give last name). He got a partial match which led to a potential suspect. They questioned the guy for like 8 hours and took a sample which didnt match. This story was included in every story on IGG for several years after it came to prominence warning of the evils of IGG.

This gives the story in detail...Spoiler the ACTUAL killer was caught with...wait for it...IGG.

en.wikipedia.org/wiki/Murder_of_Angie_Dodge#Familial_searching_crime_solving_breakthrough

Also there is a case in the US known as the Yogurt Shop Murders, you can google for the whole story but 4 girls were killed at closing time at a yogurt place in Austin Texas.There was some DNA left at the scene. They have and I am not sure why this is,probably a early attempt at IGG, a YST-R DNA sample. This was run against an FBI database generated from prisoners for research and came up with a possible match. The samples were coded by numbers and the FBI refused to give up the name. In the end they did and as you might say, no joy. YST-R only tracks the paternal side and provides much less basis for cross analysis.

In short you sometimes can get there as they did, with some luck and alot of hacking but IGG is much more efficient and effective. I know a case Parabon did recently they got some matches in GEDMatch but none were effectively close enough to make economic sense to pursue. They run the DNA on a weekly basis IIRC and lo and behold someone uploaded a sample that was close enough to feasibly work and they tracked the guy down.