back
32 comments
Related. Others?

Mystery of Somerton man’s identity solved after 73 years, researchers say - https://news.ycombinator.com/item?id=32251717 - July 2022 (14 comments)

Australia Exhumes the Somerton Man, and His 70-Year Mystery - https://news.ycombinator.com/item?id=27250937 - May 2021 (5 comments)

The Lost Man - https://news.ycombinator.com/item?id=9666581 - June 2015 (4 comments)

The Body on Somerton Beach (2011) - https://news.ycombinator.com/item?id=7913942 - June 2014 (14 comments)

I loved reading this. A nice short but true mystery solving story.

But the epilogue about how being an electronics engineer gave them a bit of a unique edge feels like a self-serving stretch. The techniques in the story seem accessible to any reasonably well-learned individual if they’re motivated enough to sleuth. All the advanced work was done by other people. And there’s a tinge of the classic, “the established discipline is living in the past but us engineers can do it better…”

I think you're misrepresenting the text a bit. It says:

> What’s interesting about solving such a case is how it relies on concepts that may seem counterintuitive to forensic biologists but are quite straightforward to an electronics engineer.

They're only claiming it's "counterintuitive" to biologists trained in other methods.

Imputation is a fairly common technique that would be familiar to most molecular or computational geneticists, especially those used to working with poor samples. There are software packages to assist in the calculation.
The point is imputation is used regularly in genetics, just not on forensics. And for good reasons. This case is old enough to be a curiosity, an amusing reading. But 70 years ago it was a potential murder case.

Imagine a forensinc lab using imputation to identify some samples and arresting someone based on them. It takes only a slighly interested lawyer to destroy that evidence at trial, saying that imputation is, literally, making up a good amount of data. In fact, defendant lawyers put a lot of effort in invalidating evidence, and imputation is quite easy to attack.

This case is a good sample to build upon it, to try to introduce imputation as a victim identification. But forensics move slowly.

> And there’s a tinge of the classic, “the established discipline is living in the past but us engineers can do it better…”

It's a classic for a reason :)

but it is belong biology region. SNPs is a old concept in genomics.
Amazing! I wonder how much this rules out the espionage, or the connection to Jo Thomson though. I wonder if she sold him the poison or something. In any case, I am morbidly fascinated by stories of people dying alone, without children, and being forgotten. Perhaps because I feel it could be my own fate. Certainly more common than any of us are prepared for, as we are all decendents of the genetic winners.
What fascinates me about this thought is that being a genetic winner in modern society could just mean you were qualified enough to donate to a sperm bank.

No more successful relationships or sinister ways of ensuring genetic survival are required.

You usually have to meet some criteria (health, education, age) in order to donate to a sperm bank. And after this selection you have to be selected which is again based on more criteria.
yeah but as a donor to a sperm bank, you can certainly be forgotten much more easily than a real parent. It's like being an uncle or something.
odd how "AI Face Recognition" sounds so authoritative, yet "machine learning Face Recognition" is also accurate but does not hide the inherent uncertainty, guesswork, noise and error in the process
AI is the new marketing term. Machine learning is old.
I grew up decades ago hearing about the Somerton Man from my father, who read books on the subject and was fascinated by it. “Somerton Man was a murdered spy” was a popular theory for quite some time and entire books were written on the subject. My father is now long gone, but I wish he was alive to learn about the mystery being solved.
> Note that SNPs have around 50 to 150 base pairs of nucleotides

What does this mean? SNPs being, by definition, single nucleotide polymorphisms, how does the author jump from 1bp to 50-150bp? Or is he referring to the surrounding DNA region?

Written like this, it doesn't make sense since a SNP has a length of 1. However, you need longer DNA fragments to be able to correctly determine the SNP, and old DNA tends to be fragmented. They are likely referring to those fragment lenghts.
“Single” in this case refers to the polymorphism not the segment length.
Could you elaborate on this? Are you claiming that SNPs do not generally refer to only one nucleotide?
According to wikipedia [1] it is a single nucleotide (or rather a single base pair). But you obviously need more base pairs to identify which base pair you are looking at in the DNA. Presumably 50-150 base pairs is the length of DNA fragments required to be able to identify where in the genome it is.

[1]: https://en.wikipedia.org/wiki/Single-nucleotide_polymorphism

> But you obviously need more base pairs to identify which base pair you are looking at in the DNA.

But that's not a SNP, then, that's a read that happens to contain a SNP.

Generally a SNP is addressed by giving the reference base, the mutated base, the chromosome, the base position in that chromosome, and the assembly used to align the read to that location. For example, a T->C mutation at chr2:25164877 using hg38.

Yes, you need to have the whole read to align to your genome assembly to detect the SNP. But before you do that, you don't know whether you have a SNP. And after you do that, the rest of the read is not useful if what you care about is the SNP.

If you know so much about the topic why ask questions about it?

From the context of the article it's pretty clear that the author was talking about the fragment sizes needed for analysis.

I didn't ask questions about the article, only about another commenter's (incorrect) claim. The original question about the article was someone else.

I have a graduate degree in bioinformatics and have worked in the industry for close to a decade, I have a strong background in the exact topic, yes.

There is another acronym, MNP, that means Multiple Nucleotide Polymorphism, but it's rarely used. It's very common to use SNP even if the variation is two or three nucleotides.
Sure, but three is a far cry from 50-150, which would almost certainly be reported as a rearrangement or some other sort of structural variation, and would be detected with an entirely different method than SNPs.

Alternately, they're talking about their sequencing reads and calling them SNPs.

I think I should just chalk this one up to my own personal Gell-Mann amnesia and let the article authors do their thing.

Answers with Joe Scott did two videos on this story, the latest one 7 months ago revealing his name.

One year ago: https://www.youtube.com/watch?v=DCDOYDNZUMI 7 months ago: https://www.youtube.com/watch?v=ve_JrmYsU0I

The article mentioned 'elliptical suicide note' what did they mean by that? It's not a code? Just random letters?
I think they mean the scrap of paper torn out of the book:

> In 1949, a pathologist had found a bit of paper concealed in one of the dead man’s pockets, and on it were printed the words Tamám Shud, the Persian for “finished.”

The scrap of paper they found in his pocket which had a scrap of paper with words equivilant to “The End” in a foreign language, ripped out of the last page of a book.
Is this using a DNA profile -> AI image? If so that is an amazing advance.
Tldr they waited until DNA sequencing and databases were good enough
But what is good enough?

Mistakes are still make quite regularly. Most people (all?) still can't grabs what "confidence interval" or "probability" are.

Not sure I understand your question