Substack cometh, and lo it is good. (Pricing)

Open Thread, 09/29/2019

Jürgen Osterhammel’s Unfabling the East is a book I really want to read someday. But I have to finish The Transformation of the World: A Global History of the Nineteenth Century first.

Ancient genomic regulatory blocks are a major source for gene deserts in vertebrates after whole genome duplications.

What Whales and Dolphins Left Behind for Life in the Ocean.

U.S. Officials Warn of Rising Threat From Qaeda Branch in Northwest Syria. Like Leftist terrorism in the 70s, this will linger for a while.

Preying on Children: The Emerging Psychology of Pedophiles. “The biological clues attached to pedophilia demonstrate that its roots are prenatal,” said James Cantor, director of the Toronto Sexuality Center. “These are not genetic; they can be traced to specific periods of development in the womb.”

“CCR5-∆32 is deleterious in the homozygous state in humans” – is it?. The paper that inspired this blog post will be retracted now, as some of the concerns expressed by the author of the post seem to have been warranted.

Adaptive Tree Proposals for Bayesian Phylogenetic Inference.

Removing reference bias in ancient DNA data analysis by mapping to a sequence variation graph.

Is Yamnaya overrated? I think a lot of the time these “ancestral” populations are just good proxies for the “true” ancestral group. That is, they are sister groups to the ancestral group. This is clear.

“The Aryans”.

ASHG 2019 meeting soon. Excited.


7 thoughts on “Open Thread, 09/29/2019

  1. Medical Genomics news:

    “Doctors Limit What to Tell Patients About Their DNA Test. Should They?: Genetic scans provide lots of information, but only a fraction is returned to patients; ‘We don’t want to frighten people’” By Melanie Evans and Anna Wilde Mathews on Oct. 4, 2019

    The Mayo Clinic is scanning 20,000 genes for thousands of patients to study genes’ role in disease. It will hand over results for just 59.

    Mayo will look for certain disease-causing gene variants for heart disease or breast cancer, and offer results to patients who have them. But it doesn’t look at variants for early-onset Alzheimer’s or Lou Gehrig’s disease, meaning patients will remain in the dark. … If told, some patients could seek unnecessary or harmful care, or unduly worry about a disease they may not get.

    * * *

    Some doctors, however, say it is paternalistic to withhold information if patients want it. Several studies have suggested that most patients want to learn their own genetic results. “It’s their body and their DNA. We have a responsibility to scientific truth and clear communication,” including helping patients integrate results into their care, said Dr. Robert Green, a geneticist who is a professor at Harvard Medical School. He said doctors shouldn’t “take the Jack Nicholson line, ‘You can’t handle the truth!’ ”

  2. Sounds like a GWAS study on drug addiction. Not something you would expect from the Republican Attorney General of Ohio.

    “Attorney General Dave Yost seeks DNA markers underlying opioid addiction” By Randy Ludlow in The Columbus Dispatch on Oct 3, 2019

    “Yost is using $1.6 million in settlement funds from various lawsuits to underwrite a pair or projects to address Ohio’s opioid addiction and death crisis, which only last year began to show signs of easing.

    The genetic marker study will seek DNA cheek-swab samples from up to 1,500 emergency room patient volunteers at Ohio State University’s Wexner Medical Center and the University of Cincinnati Medical Center. Patients confidentially will be asked about factors associated with addiction and their use of opioids, with DNA testing to look for 180 genetic markers suspected of association with opioid addiction.

    Scientists from Ohio State, the University of Cincinnati, Bowling Green State University and Case Western Reserve University will work with a pair of contractors to analyze the data. Dr. Jon Sprague, director of science and research at Yost’s office, and Dr. Caroline Freiermuth of the University of Cincinnati will lead the work.

  3. Book question. I am trying to do at least a minimum purge of my book collection which has overrun available shelf space. I found the following which I have (sadly) not gotten around to reading:

    “Great Human Diasporas” Cavalli-Sforza 1995

    “Saxons, Vikings, and Celts” Sykes 2006

    “Before Dawn” Wade 2006

    “Grandfather’s Tale” Lessard 2012

    Are they all outdated? Should I keep any of them?

  4. @Razib, btw, I remember you blogging on this preprint on Dystruct algorithm back in 2018 (this one:, blogged by you at

    Well, randomly found while browsing that it looks to have been published in August 2019, and with some additional analyses and extended ancient dataset:“Inference of Population Structure from Time-Series Genotype Data”

    May be worth a look. Their added Figure 8 which splits ancestry in modern populations between components representing CHG, Iran_N, EuroHG, EEF, and Levant_N, contrasted with an ADMIXTURE which conflates CHG and Iran_N, and EEF and Levant_N is particularly interesting.

    Even more might come out of this algorithm with a further extended West Eurasian dataset than they use (they use one based on Lazaridis 2016, which is actually pretty old hat now, although a vast improvement on the Haak 2015 set they had in the preprint!), and with a richer set of non-West Eurasian adna that has begun to be available.

  5. Supplement from the DyStruct paper above breaks down the specific components and populations being dealt with in their Fig 8 (supplement:


    The K10 DyStruct has fewer splits within Africans (a consequence of little African adna), but presents European farmers as ancestral to modern Europeans, in contrast to ADMIXTURE which seems to present an implausible result, with the Levant component ancestral to present day Europeans.

    (DyStruct instead presents Levant component as present only in some Europeans, primarily Greece, Cyprus, Spain and Sicily, petering out with distance such that there is very little in Ukraine and North Russia and Central France, and none at all in Scotland, England, Belarus and Lithuania).

    Main problem with using DyStruct that’s obvious here might be prohibitive time 42.55 hours from K=10 on Lazaridis 2016 dataset, would probably run to be infeasible on 2019 datasets.

Comments are closed.