4 ms·

> They aren't truly geographical- they're based on population structure in the genome. Determining whether people are genetically similar is indeed well-studie

by smt88 2y ago

> They aren't truly geographical- they're based on population structure in the genome. Determining whether people are genetically similar is indeed well-studied. The problem is saying with certainty that you know where their ancestors lived. There are a few ways the results end up being inaccurate, and they apply to all the companies. The first and most important is that the data is from where people live now or in the very recent past (as I alluded earlier). That means they can only tell you how closely-related you are to people in a certain geographic area, whether those people have been there for 100 years or 1,000 years. The second is that the data sets are filled with mostly European data, so anyone with even a bit of non-European ancestry is going to have less-accurate results[1] presented to them with a misleading degree of confidence. The third is that there isn't an objectively "correct" algorithm for determining if someone is from one place or another. Every company has their own, which is why people get (sometimes wildly) different results from different companies.[2][3] And the fourth is that the data collection doesn't seem to be especially accurate, perhaps related to the quantity or variety of SNPs that the testing company collects data about. > Above you said "ex has a doctorate in genetics". Correct. And the same way I take climate scientists' word for it that climate change is happening, instead of getting 10 degrees in different fields and gathering data myself, I understood her explanation and trust her expertise. Being a science consumer is always about trusting other people who have done the research, not about learning it all yourself and gathering your own data. If you can find someone with a doctorate in genetics who has no affiliation with testing companies and believes these tests accurately tell someone where their ancestors were from, I'd be happy to consider the counterpoint. I have never seen such a thing in my research. Regardless, all of the commercial operations have given twins differing results, so we can be fairly certain that any accurate way of determining genealogy from DNA has not been commercialized. These tests, however, are good for finding relatives or determining how closely-related two people are. 1. https://www.nih.gov/news-events/previous-genetic-association-studies-involving-people-european-ancestry-may-be-inaccurate https://www.nih.gov/news-events/previous-genetic-association... 2. https://www.livescience.com/63997-dna-ancestry-test-results-explained.html https://www.livescience.com/63997-dna-ancestry-test-results-... 3. https://www.vox.com/science-and-health/2019/1/28/18194560/ancestry-dna-23-me-myheritage-science-explainer https://www.vox.com/science-and-health/2019/1/28/18194560/an...

dekhn 2y ago
My doctorate is in biophysics with deep experience in genomics. I've attended talks from the world's leading experts (as well as scientists/researchers/engineers) who work in this field, and read papers. From my inspection of the data, identity by descent does exactly what the researchers say, with a level of accuracy that is consistent with our current understanding of population structure and its changes in the past. BTW, the main reason african genomic inference is so complicated is not the lack of genomic data, but rather that africa itself has far higher genetic diversity than any other continent, due to populations (whose genomes we can infer!) moving and interbreeding for a long time before humans left africa and spread out to the rest of the world. (Article #1 is correct, there are definitely populations for which our current models are inaccurate due to bias and underlying complexity. I am not disagreeing with any of that). (Article #2, where an individual compared the results, shows that the results are nearly identical, and well within the expected accuracy of the tests) (Article #3 isn't interesting- yes, 23&Me and other companies have marketed truly terrible ideas. That doesn't mean IBD is scientifically invalid. The article also strawmans many of the realities underlying the nature of the tests and their relationship to disease prediction, or any other phenotype of interest).
- smt88 2y ago
  We seem to be talking past each other. You're saying it's possible to accurately determine how closely related individuals (and even groups) are. I agree. But you can't determine where those individuals' ancestors lived at any given time unless you also have historical data. Land does not imprint upon DNA. So these companies can tell me that I have many distant cousins in, say, Brazil and Portugal, and we can make an inference that I'm of Portuguese descent. But clear-cut situations like that are extremely rare unless good genealogy records are kept as well and can be combined with genetic testing.
  - dekhn 2y ago
    Land definitely imprints on DNA (in genomic time intervals). I don't think people using IBD are claiming they can explain the fine structure of population movements. I didn't claim it, it seems unique to your contributions to this thread.

by smt88 2y ago

dekhn 2y ago

My doctorate is in biophysics with deep experience in genomics. I've attended talks from the world's leading experts (as well as scientists/researchers/engineers) who work in this field, and read papers. From my inspection of the data, identity by descent does exactly what the researchers say, with a level of accuracy that is consistent with our current understanding of population structure and its changes in the past. BTW, the main reason african genomic inference is so complicated is not the lack of genomic data, but rather that africa itself has far higher genetic diversity than any other continent, due to populations (whose genomes we can infer!) moving and interbreeding for a long time before humans left africa and spread out to the rest of the world. (Article #1 is correct, there are definitely populations for which our current models are inaccurate due to bias and underlying complexity. I am not disagreeing with any of that). (Article #2, where an individual compared the results, shows that the results are nearly identical, and well within the expected accuracy of the tests) (Article #3 isn't interesting- yes, 23&Me and other companies have marketed truly terrible ideas. That doesn't mean IBD is scientifically invalid. The article also strawmans many of the realities underlying the nature of the tests and their relationship to disease prediction, or any other phenotype of interest).

smt88 2y ago
We seem to be talking past each other. You're saying it's possible to accurately determine how closely related individuals (and even groups) are. I agree. But you can't determine where those individuals' ancestors lived at any given time unless you also have historical data. Land does not imprint upon DNA. So these companies can tell me that I have many distant cousins in, say, Brazil and Portugal, and we can make an inference that I'm of Portuguese descent. But clear-cut situations like that are extremely rare unless good genealogy records are kept as well and can be combined with genetic testing.
- dekhn 2y ago
  Land definitely imprints on DNA (in genomic time intervals). I don't think people using IBD are claiming they can explain the fine structure of population movements. I didn't claim it, it seems unique to your contributions to this thread.