An argument over sex that has been going on for more than a year is finally seeing the light of day. Today, scientists at the University of Cambridge, UK, and Harvard Medical School in Boston, Massachusetts, let the world in on a long-running discussion over whether or not humans and Neanderthals really interbred — and how you go about proving it.
I’ll get to the sex. But this debate underscores a topic I wrote about last month (see ‘Geneticists eye the potential of ArXiv‘) that noted that high-profile papers from population geneticists are beginning to appear on the preprint server, once the domain just of theoretical physicists. That story is relevant because a new paper, entitled ‘The date of interbreeding between Neandertals and modern humans’, was posted to ArXiv on Friday. Meanwhile, a second paper raising doubts about human-Neanderthal hanky-panky appears in the Proceedings of the National Academy of Sciences (PNAS) today.
Both papers were presented at conferences more than a year ago. Their publication today raises the question of whether this debate would have been more timely if it had occurred on preprint servers such as ArXiv.org and not at specialist conferences and behind the walls of peer review.
In putting a date — 37,000–86,000 years ago — on human–Neanderthal relations, Harvard’s David Reich attempts to address a question created when he and his co-author Svante Pääbo, of the Max Planck Institute of Evolutionary Anthropology in Leipzig, Germany, published a draft sequence of the Neanderthal genome in 2010.
Just about every science reporter (see ‘European and Asian genomes have traces of Neanderthal‘, for example) led with the conclusion that Neanderthals and non-African humans had interbred.
But Pääbo, Reich and their co-authors said that there could be another explanation for their observation that all non-Africans surveyed (that is people with deep Asian and European ancestry) owed about 1–4% of their genome to Neanderthals, while African genomes contain no detectable Neanderthal DNA. Neanderthals and humans share an ancestor that lived in Africa about half a million years ago. It is possible that modern non-African’s Neanderthal DNA dates to the time of these early humans — and not more recent escapades. By chance or quirk of geography, the humans who left Africa could have been more closely related to Neanderthals than the humans who stayed behind.
In other words, it only looks like humans and Neanderthals interbred when we compare their genomes, but 500,000-year-old population structure in Africa is the real explanation. Anders Eriksson and Andrea Manica, at Cambridge, elaborate on this scenario in their new PNAS paper.
Not so, say Reich and Pääbo in their ArXiv paper, which reports a nifty new method for determining when two populations interbred. It’s based on the fact that our maternal and paternal chromosomes reshuffle after each generation. This mixing makes the contiguous chunks smaller with each generation. Reich and his colleague Sriram Sankararaman take advantage of this feature to conclude that humans and Neanderthals interbred between 37,000 and 86,000 years ago, and probably more like 47,000–65,000 years ago.
Reich tells Nature that the paper will soon be published in PLoS Genetics. Meanwhile, Monty Slatkin’s team at the University of California, Berkeley, came to a similar conclusion to Reich’s using a different method. They published their results in April and presented them at a conference last summer.
Fair enough, one might say, this is how science works. One paper raises questions that are addressed by others. But Reich believes that the discussion would have been different if it had happened in the open. The PNAS paper questioning the Neanderthal admixture addresses issues swirling around two years ago, but not Reich and Slatkin’s latest work. “It’s been an issue for several years. They were right to work on this,” says Reich. But now, “it’s kind of an obsolete paper,” he says.
Manica says that more data are needed to reach any firm conclusion on human–Neanderthal relations. “I think one take-home message is that establishing the presence and strength of hybridization is far from trivial if you only have one Neanderthal genome.” He supports pre-publication, in theory. “The problem is that papers are not routinely uploaded as soon as they are ready, as it the case in physics. So, for ArXiv to be functional, everyone needs to upload their drafts on a regular basis.”
Reich, who posted another high-profile paper on the genetic history of southern Africans to ArXiv two weeks ago, thinks this debate would have been different if geneticists routinely posted to preprint servers. The conversation could be happening in near real-time and not with a two-year lag. “We think there’s no reason not to [post to ArXiv] and it’s interesting for other people to read about that work,” Reich says. “Maybe it would have been helpful in this context as well.”
Updated 23:21 to include comment from Manica.