Thursday, 21 November 2013

Day 1 at the Royal Society's 2013 Ancient DNA meeting

I spent two very interesting days this week attending the Royal Society’s meeting on Ancient DNA: the first three decades. Recorded audio of the presentations will be available on the Royal Society’s website at some point and the papers will be published in a future issue of Philosophical Transactions B. While at the meeting I made notes during the talks, and I thought that until the recordings have been uploaded to the website these notes might be of interest to those who were unable to attend the meeting. These notes are not intended to provide comprehensive coverage, and I only jotted down items that I personally found of particular interest. My primary focus is on the genealogical applications of DNA testing, and my interests will, therefore not necessarily coincide with those of other researchers. Many of the technical and scientific details of the talks were well outside my expertise. The accuracy of my notes and my interpretation of the lectures is not guaranteed, but I hope that some people might find the information useful.
The Royal Society in Carlton House Terrace, London SW1 - 
the venue for the Ancient DNA meeting.

Full details of the meeting, along with speaker biographies, can be found on the Royal Society’s website. The abstracts for these talks have not been made available on the website though they are all included in the programme which was issued to attendees.

A related satellite meeting is taking place in Buckinghamshire and finishing tomorrow. The speaker’s biographies and the abstracts are available on the website for the this meeting. I was not able to attend this event but I hope that other attendees will provide reports in due course.

Erika Hagelberg, University of Oslo, Norway
Ancient DNA: the first three decades
The first article on ancient DNA was published in 1984. It was a report of the cloning of a small piece of DNA from the skin of an extinct equid (a member of the horse family) that had been preserved in a museum.
The second important ancient DNA paper was on molecular Egyptology.
A lot of the early research centred on Allan Wilson’s lab
In the early days ancient DNA testing was done on the workbench without any protective clothing.
PCR [polymerase chain reaction – a process for amplifying DNA] was introduced in the late 1980s.
The first PCR machine was made with a kettle.
The late 1990s saw the development of standards of authenticity. Hagelberg felt that the new standards stifled research and open discussion.
The big technological advances in recent years have been in bioinformatics, contamination filters and next generation sequencing.
The early studies on ancient DNA (magnolia leaf, an insect embedded in amber) are now not considered very credible. It is also difficult to reproduce these early studies.
The first ancient DNA newsletter was published in 1992.
The limit for ancient DNA was originally thought to be 5000 years.
1 March 1990 Angel of Death newspaper article on the DNA of Mengele. This was the first use of DNA in forensics.
The 1990s also saw the DNA analysis of the remains of the Russian Imperial family. Some people disputed the results.
1994 Dinosaur DNA turned out to be human DNA
1997 Ryk Ward and Chris Stringer publish a paper in Nature in which they outline standards for ancient DNA research
2000 Cooper and Poiner letter in Science. “Do it right or not at all”
Hagelerg said that this was often interpreted as “Do it with me or not at all”.

Christine Keyser, University of Strasbourg, France
Past human populations in Eurasia
Keyser reported on an ancient DNA study of samples obtained from 150 graves in Yakutia  in Northern Siberia.
146 bodies were found. They were frozen at the time of discovery. Genetic data was obtained from 130 bodies.
Optimal ancient DNA is obtained from bone.
Smallpox found in Yakut graves – identified by histology.
Y-chromosome analysis was done using a Y-filer kit (17 Y-STRs). There were 20 different haplotypes. A strong founder effect was found with one haplotype shared by 29 males (46%). They went up to 23 STRs on these samples but found only three differences in the 29 males.
For the mtDNA analysis they tested HVR1 and the coding region. There were 44 different mtDNA haplotypes (n=130) with haplogroups C and D predominating.
IrisPlex and HirisPlex were used to determine hair and eye colour. Six SNPs used to detect eye colour. Brown hair and brown eyes.
SNP testing. N1c1 was the predominant Y-DNA subclade.
Full mtDNA genomes sequenced. D5a2a most common subclade.

Anne Stone, Arizona State University, USA
Impacts of colonisation in the Americas
Anne Stone was invited to speak at the last minute after the scheduled speaker, Ripan Malhi, had to withdraw. Malhi’s talk was to be on the subject of “The evolutionary history of Native Americans”. There is a summary of his planned talk on Science Daily in an article entitled Ancient, modern DNA tell story of first humans in the Americas.

Stone's talk focused on the impacts of colonisation in the Americas.
The initial colonisation of America took place between 18,000 and 25,000 years ago.
The post-Clovis theory of colonisation is dead.
The major part of Stone’s talk focused on the Salesia mission in Tierra del Fuego.
TB was the leading cause of death at the mission. No genetic evidence of TB found in her study.
Targeted enrichment to get full mt genome.
The genetic evidence shows that TB was already in animals in America before humans arrived.
Hershberg et al 2008 paper on the biogeography of M.tuberculosis.
The genetic testing of Native Americans depends on view of individual tribal groups.

Questions from the audience
Q What is the evidence for the pre-Clovis theory?
A The genetic evidence for pre-Clovis colonisation of America is based on signals of expansion. Human coprolite data is also pre-Clovis [coprolite = fossilised poo!].

Helena Malmström, Uppsala University, Sweden
The Neolithic transition in Scandinavia
Farming started 12,000 years ago in the Near East and 7,000 years ago in Northern Europe.
In Scandinavia hunter gatherers and farmers co-existed for a period of about 1000 years.
The hunter gatherers (Pitted Ware complex) and the farmers (Funnel Beaker complex) had different maternal lineages.
Haplogroup U was found at the highest frequency with U4 top of the list.
Autosomal SNP analysis showed that the Neolithic hunter gatherers differ from modern Europeans and were most like Sardinians and Basques.
[DK note: For background see the 2012 Nature News article by Henry Nichols Ancient Swedish farmer came from the Mediterranean and the 2009 paper by Malmström et al.] 

Carles Lalueza-Fox, Institute of Evolutionary Biology (CSIC-UPF), Spain
Neandertal paleogenomics and the El Sidrón cave
This was an excellent and sometimes humorous talk on the exciting findings from El Sidrón cave in Asturias, Spain.
Lalueza-Fox started by sharing a number of illustrations showing how our perception of Neanderthals has changed over time. We now know that they used language, and they lived in family and social groups. The final picture representing the current thinking showed a picture of a Neanderthal mother and child looking not much different from modern humans.
See also the modern reconstruction picture shared by @mjpallen on Twitter.
 Laleuza-Fox took us on a photographic tour of El Sidrón cave. A group of Neanderthal individuals were found in this cave. They had been trapped in the cave after a rock fall and their DNA provides a snapshot in time of a Neanderthal social group.
Complete mtDNA genomes were obtained.  Three different Neanderthal mtDNA haplogroups were found which Laleula-Fox has labelled A B and C. 7/12 were A. 1/12 was B and 4/12 were C. Three adult males had the same mtDNA but the three adult females had different mtDNA. This is indicative of patrilocal reproductive behaviour.
There were cut marks on all the remains – evidence of cannibalism.
Laleuza-Fox et al 2007 paper in Science. Some Neanderthals had red hair

David Reich, Harvard Medical School, USA
Insights into population history from high coverage Neandertal and Denisova genomes
[DK comment: Why do Americans spell Neandertal without an H but pronounce the word as though it does have an H. Why do Brits spell Neanderthal with an H but pronounce it as though it doesn’t have an H?]
This was the highlight of the first day’s talks. It was delivered at breathtaking speed, barely allowing us time to digest the content on the slides. I would have liked to have had a pause button so that I could stop and look at everything again in more detail.
Neanderthal gene flow is about 2%:
1.72% in Europeans
1.89% in East Asians
(Confidence intervals were provided but the slide disappeared to quickly for me to note them.)
Autosomal DNA analysis used a recombination rate of 10cM per 10 generations, 100 cMs per 100 generations. I spotted Graham Coop’s name on this slide but wasn’t sure whether Reich was citing the paper The geography of recent ancestry across Europe 
We now have Neanderthal sequences from three different locations: Croatia, Russia and the Altai Cave in the Altai Mountains in Siberia. This is the cave where Denisovan DNA was found but the latest analyses show that Neanderthals also lived there.
Archaic split 77-114 kya.
There were multiple gene flows.
In the original Denisovan study DNA was extracted from the little finger of a young girl. The samples date back more than 50,000 years. DNA has now also been extracted from a molar.
1.9 fold coverage of genome.
Denisovans are more closedly related to Neanderthals than to humans. Their mtDNA is twice as deep compared to Neanderthals than humans.
Denisovans are closely related to people from New Guinea. New Guineans have 4.6% Denisovan and in addition 2.5% Neanderthal.
2013 paper to be published on Altai Neanderthal found in same cave. Sequencing done at high resolution 52x coverage.
The archaic populations have a very low level of genetic diversity. The Altai Neanderthal are highly inbred.
Reich showed us a number of slides exploring a number of hypotheses he investigated on the relatedness of Denisovans to Neanderthals and humans. He concluded that “Denisovans harbour ancestry from an unknown archaic population unrelated to Neanderthals and modern humans”.
[DK note: This finding was anticipated by Graham Coop in his Haldane’s sieve blog post Thoughts on: The date of interbreeding between Neandertals and modern humans.]
New research has shown that Denisovan DNA is now found in East Asians. See the Cooper and Stringer 2013 paper: Paleontology. Did the Denisovans cross Wallace's Line?
Conclusion: gene flow between diverged humans was common in late Pleistocene and there were five events.

Questions from the audience
Q Does this mean humans copulated with Neanderthals? A Yes!
Q Does this mean humans fancied Neanderthals? A Yes!

Reich’s talk seemed to be the one that was attracting all the interest from the media. Ewen Callaway, the reporter from Nature, was at the conference and he has already written an article for Nature Breaking News which can be found here. There is further coverage from Michael Marshall in New Scientist.

[DK note: The abstract for this paper also mentions Neanderthal X-chromosome ancestry. I don't know if I missed the mention of the X-chromosome in this high-velocity presentation or if it was perhaps not covered. Here is the relevant extract from the abstract: "The average Neandertal ancestry on the X chromosome is about a fifth of that in the rest of the genome. It is known from studies of many species that genetic variations causing hybrid sterility concentrate on chromosome X. This is consistent with Neandertals and modern humans having been on the edge of biological incompatibility when they met and mixed.]

Johannes Krause, University of Tübingen
Ancient pathogen genomics: what we learn from historical diseases
The Black Death killed 30-50% of the population of Europe. It probably originated in China. Yersinia pestis has the biggest diversity in China.
99% of pestis genome sequenced at 30x coverage.
Yersinia pestis MRCA within last 4000 years.
There is nothing in the genome to explain the high mortality rate.

Christina Warinner, University of Oklahoma, USA
A new era in paleomicrobiology: microbiomes
If you go by the number of cells in our body we are 90% bacteria.
The bacteria in our bodies weigh around three pounds.
The bacterial genome is also known as the accessory genome.
There has been a 38-fold increase in the number of known bacteria in the last seven years.
Best estimate before NGS is 500 species of bacteria in mouth. After NGS, 19,000!
You can get lots of DNA from calculus.

[DK note: I'm afraid I was flagging at this point after a 5.15 am start to my day and only four hours' sleep. This talk was highly technical and much of it was over my head. The take-home message from the final talk was that this is an important emerging new field for the study of ancient DNA.]

The recordings of all the lectures from this meeting are now freely available on the Royal Society's website.

See also
My notes from Day 2 at the Royal Society's 2013 Ancient DNA Meeting

© 2013 Debbie Kennett


Unknown said...

Great summation, Debbie. I would love to have been able to attend. Wish they had filmed the presentations! Thank you for taking the time to take notes and publish what you learned

Africa Gomez said...

Thank you for the superb review! I wish I was there, although it is a fun topic for me, not my research topic.
You might want to correct typo:
"Farming started 12,000 years ago in the Near East and 70,000 years ago in Northern Europe"
Surely you mean 7,000

Could this be the Fagundes paper:



Debbie Kennett said...

Thank you Nora. I'm glad the report was of interest. You will be able to listen to the audio though it's not the same as seeing a video or, even better, being there in person.

Thank you Africa. I did mean 7,000 years ago for the introduction of farming. Such a dramatic shift in the date would have been headline news!

Thanks for the link to the Fagundes paper. I'd mis-spelt the name as Fagunes which is why I couldn't find it.

Brian Swann said...

David Reich from Harvard Medical School was very impressive. His talk at the satellite meeting was to try and define better how much Neandertal could have entered the human genome, how much remains and how much Neandertal DNA was replaced fairly quickly after interbreeding. His conclusions were that the last item, loss of Neandertal DNA was appreciable across many sections of the genome. One of the principal mechanisms for that was possibly an effect on the X-chromosome of the Neandertal-human hybrid leading to issues with further breeding success down the genetic line and either replacement of the introduced Neandertal genes or decline in reproductive success. He also said, more or less, you can forget any percentages of Neandertal or Denisovan DNA admixture given by the DNA testing companies. They are almost random background noise in terms of percentages. His new data yielded much tighter error bars and lower percentages of Neandertal incorporation into humans. As he works in collaboration with Svante Paabo's group, his statements have a high credibility.

Debbie Kennett said...

Thanks Brian for sharing that. From the tweets I read Reich estimates that about 40% of Neanderthal DNA can be found in humans but each individual only has a tiny percentage.

Geno 2.0 has been finding "Denisovan" DNA in lots of Europeans (it's only supposed to be found in Melanesians and Polynesisans and now East Asians). I asked Michael Hammer about this at WDYTYA and he said that they must be wrong. In any case the Denisovan estimate was only ever experimental and it now appears that the Genographic might be removing these estimates altogether. See this very interesting interview here with Spencer Wells:

I've copied the relevant sections below:

"GB: Is the scientific community, specifically the genetics community, in consensus on what to look for in terms of neanderthal or denisovan DNA in modern humans?

SW: Oh, god, no. [Laughs] It’s all very much evolving, and we caught wind of some interesting new stuff at a genome conference held at Cold Spring Harbor back in May where Svante Pääbo — who is the world expert in neanderthal and denisovan genomics; his lab is the one who sequenced them — made an announcement that with the new denisovan genome they have they’re now reevaluating what a denisovan is because it seems to be about 20% neanderthal. And the fact is that we don’t really know what these things are, and the extent to which there is overlap and interbreeding between neanderthals and denisovans.

So, the denisovan calculation in particular is extremely experimental, and we say that on the website. And in fact, I think the story is going to get so murky about what a denisovan actually is, we’re probably going to stop reporting that because it’s simply going to be uninterpretable. Everybody has some ancestral admixture with these hominids. The neanderthal estimate is probably pretty valid — the denisovan one, I think at the moment, is just totally up in the air — up for grabs. And that paper hasn’t come out yet. It’s supposed to come out at some point this year, and when it does, we’re going to reevaluate whether or not we’re going to continue to test for denisovan DNA."

With regards to estimates of Neanderthal DNA from commercial tests it seems that this finding is more secure. See Eric Durand's white paper for 23andMe:

Eric Durand worked on the project to decipher the Neanderthal genome.