Nature Podcast 6 July 2006
Introduction
This is a transcript of the 22 June edition of the weekly Nature Podcast. Audio files for the current show and archive episodes can be accessed from the Nature Podcast index page (http://www.nature.com/nature/podcast), which also contains details on how to subscribe to the Nature Podcast for FREE, and has troubleshooting top-tips. Send us your feedback to mailto:podcast@nature.com
Advertisement
With Bio-Rad's new XPR protein interaction technology you can characterize up to 36 real time interactions on a single chip. On the web at http://www.biorad.com/proteininteraction.
Chris Smith: This week, how the brain recognises faces and it turns out that what they say about foreigners is entirely true.
David Leopold: Ironically it's possible that these recordings in monkeys might inform us about why when we visit a foreign country, all of the people look the same to us in the beginning.
Chris Smith: We'll also be hearing how the US could be about to put it's faith in a new generation of nuclear weapons, but these ones come with no guarantees.
Geoff Brumfiel: It's programmed to look for a new kind of warhead. They want to build this warhead without testing it.
Chris Smith: And after dire predictions from the San Andreas Fault two weeks ago, scientists have now discovered a new form of earthquake, although you'd be forgiven for missing this one.
Peter Savelli: On January 27th 2005, a large earthquake struck beneath one of America's largest national parks, yet nobody felt the shaking.
Chris Smith: Hello. I'm Chris Smith. Welcome to this week's Nature Podcast. Now, first up this week, faces... things that the brain clearly holds in high regard because there's a specific region devoted to processing them and recognising different individuals. But how does it work out who's who, and can we be forgiven for thinking that foreigners all look the same? Well, it turns out that it's all down to a process of deduction, that is the brain deducts the face it's seeing from an internal model of what the average face should look like. Here's David Leopold. Nature advance online publication 5 July 2006
David Leopold: Face recognition is something we're obviously very good at, but interestingly we have a very poor insight as to how we recognise a face. So, we have been using a set of photorealistic faces and this is a set of faces that originates from scans of humans, three dimensional very realistic scans, but having about 200 faces scanned in this way, what it's allowed is to create an artificial space where you can create practically any face and by creating different faces, you can test specific hypotheses coming from the psychology literature about how we recognise faces, say, of intermediate identity.
Chris Smith: So talk us through the actual series of experiments you've done using this tool.
David Leopold: So the experiments consisted of showing faces to a monkey and we were able to record from a single cell in the brain of that monkey... so the brain, the cerebral cortex has a 100 billion cells and by using a very small electrode we can sneak up right next to one of them and monitor its activity whilst the monkey sees different faces. Now this kind of experiment is actually rather common and has been done for several decades. The difference in our experiment was that we were able to test these theories of face perception by showing these morphed photorealistic faces at different levels of identity and asked the monkey which face he saw at any point in time, and then correlated the face identity with the response in the inferotemporal cortex of the brain where we were recording.
Chris Smith: So, what you've essentially got is a population of cells in the inferotemporal cortex that are tuned to certain facial characteristics and the more of them that fire off, you slowly build a picture of a person's identify, I suppose?
David Leopold: In a sense. One of the surprising aspects of the current study was that the tuning of the cells in this area seemed to be particularly centred around the average face, so as one moved to increasing identity levels in different directions of this theoretical face space, in other words, to different individual faces, there was a more or less linear, at least monotonic increase in the firing of cells at increasing levels of identity.
Chris Smith: In other words, you subtract the average face from the face you're presented with, to give you the algorithm for who someone is?
David Leopold: Exactly. And it appears that this is done in a very systematic way and by examining this, we think we're beginning now to understand some of the principles by which the brain is able to so efficiently and quickly recognise faces.
Chris Smith: Now would your model, then, give credence to the claim which, to be honest, I personally think it might be true, which is that if you look at foreigners, they sometimes always look the same to you and we to them? Do you think that's true then?
David Leopold: Well, certainly there's no such thing as an absolute average face. You can certainly imagine different races, different cultures, would have a different average. So, one of the interesting aspects of this study, and certainly the direction we would like to go now, is to see what are the variables that are forming this average internal representation and how can that shift with experience? For example, if one were to go to China, it may be that in the beginning, and this has been tested actually, in the beginning your recognition of discriminating between different Chinese faces is quite poor, but after time you become quite good at it, and this may indeed involve changing neurophysiologically the stored average face that you have in your head.
Chris Smith: The NIH's David Leopold. In fact, when he was explaining his results to me, David used Prince Charle's ears as an example of a feature that would set someone apart from the average.Now from brains and faces to an Australian icon, the koala. And as if it wasn't threatened enough, a new cancer-causing virus appears to have adopted it as its host. On the trail of this new agent, and from the University of Queensland, here's Paul Young. Nature 442, 79–81 (6 July 2006)
Paul Young: I guess it started a number of years ago, about four or five years ago, when a veterinary scientist from a theme park here in Queensland, Australia, came to us with a small project where he had found that a virus, an endogenous retrovirus, was present in koalas and he thought that it may be associated with the high rate of cancers in those animals. And he asked us to pursue that line of research.
Chris Smith: And what did you find?
Paul Young: We found that, indeed, this virus was present in koalas. The surprise came for us when we tested a large number of koalas and they all showed the presence of this particular virus. So we set out to try and make the statistical link between presence of virus and incidence of cancer, and we did show that. We showed that there was a correlation between high levels of virus and the incidence of lymphoma and leukaemia in these animals.
Chris Smith: But what actually is the virus?
Paul Young: Well, the virus is a retrovirus. These are viruses that have as their genome a piece of ribonucleic acid and they have the unique ability to be able to reversetranscribe that genome into DNA and integrate it into the host chromosomes of the cells that they infect. One of the interesting aspects of retroviruses is that quite surprisingly when the human genome sequence was published about four or five years ago now, the surprise was that upwards of 8% of our entire genome was actually retroviral elements, indicating that throughout our long history, we've had many incursions of these viruses into our own host chromosomes and they've settled down into our genome and have been passed on from generation to generation.
Chris Smith: So what's different about the koalas then?
Paul Young: Well, what appears to have happened with the koalas is that we're looking at a recent infection. In the case of humans, the viruses infected us some millions of years ago, and what we're looking at in our genome are the relicts of those viral infections. With the koala, there were populations in island communities off the south of Australia that didn't have the virus at all, and in the state of Victoria, there were populations with mixed infection status. So what we appear to be seeing here is, this is a virus that has probably entered the population only in the last hundred years, and I guess the most exciting thing about that is the fact that no-one has been able to look at this event of an interaction between this sort of virus and its host genome before. How it's actually transmitting between the animals is still a mystery for us.
Chris Smith: So I guess that and the actual origin of this thing must be the next thing that you're going to look at?
Paul Young: Now that's a good question. When the virus was sequenced, the strange thing was that its closest relative turned out to be the gibbon-ape leukaemia virus. This is another retrovirus that was discovered in gibbons. Clearly, if you think about it, there's not a lot of ecological niches where koalas and gibbons cohabit, so our current working hypothesis is that infection of both koalas and gibbons has come from a third species, and we're currently on the hunt for that third species. There is a suggestion that it may be the rodent.
Chris Smith: Do you think it's going to be possible to vaccinate these animals?
Paul Young: We're actually looking at that. One of the management strategies following the identification of populations of animals that don't have the virus is to develop tests that allows us to identify those virus-free animals and introduce breeding programmes that hopefully can expand the virus-free population. But obviously what you want to do then, in the context of a wider population that's infected with this virus, is prepare a vaccine that can protect them from subsequent infection, and we are working on that at the moment.
Chris Smith: Paul Young on the trail of a gibbon retrovirus that appears to have swung its way into the koala population. Shortly, we'll be hearing how the US could be about to launch into a new phase of nuclear warhead development, but without testing the results to see if they work, a new way to spot Earth-sized planets orbiting distant stars, and silent earthquakes - quakes but without the shakes.First, though, with a look at some of this week's other news and talking to Anna Lacey, here's Nature's Jo Marchant...
Jo Marchant: Thanks. The first story this week is actually about a social scientist who has spent 30 years studying the physicists who work on gravitational waves. He wanted to test whether or not he could actually pass himself off as a bona fide physicist, so he set up a test where he had to answer various questions about gravitational wave physics. He gave answers to these questions, he had a real physicist give answers to those questions, and then got other physicists to judge which was the real physicist. Nature 442, 8 (6 July 2006)
Anna Lacey: And what was the result?
Jo Marchant: The result was that he fooled them. Out of nine physicists who were asked to tell the difference, seven of them weren't sure which was the real physicist and two actually chose the social scientist, who is Harry Collins at Cardiff University.
Anna Lacey: So why on Earth did he want to do this in the first place?
Jo Marchant: Well, as I said he spent 30 years studying these physicists and basically the nature of what they do. He is a social scientist after all. But I think one of the criticisms that is often made, and especially from scientists, is that a social scientist can never really understand the field that they studying well enough to make valid conclusions, and they rarely understand the science that's being done and what the scientists are aiming to do if they're not actually a scientist themselves. So this is Harry Collins's attempt to show that, yes, actually he can.
Anna Lacey: How long do you have to be immersed in some kind of science community in order to achieve this?
Jo Marchant: I think it's difficult to give a very black and white answer to that question. Collins has said that he's shown that three decades is certainly enough, but he also says that this kind of expertise should also be useful with, for example, anthropologists going to a remote community and wanting to study their culture, and there has been a criticism in the past that how could they possibly understand enough of what that culture is about to say anything about it that's useful. And I think the implication here is that you can gain an understanding of what a culture is talking about without actually becoming part of that culture. You can look at something from the outside and comment on it.
Anna Lacey: Thanks, Jo. We're now taking a look at how to make it to the top in the world of science blogging. Here's Nature's Declan Butler. Nature 442, 9 (6 July 2006)
Declan Butler: Hello, yes. We've been very interested in following the progress of scientific blogs which have been very few and far between compared to many other areas. So this week, we decided to try and look at what are the top science blogs among the 46.7 million blogs that are indexed by the Technorati blog search engine.
Anna Lacey: So, with all those millions of blogs, how on Earth did you go about actually ranking them?
Declan Butler: We took a selection of science blogs by looking at the number one of all the science blogs that I knew, plus all those that were listed on different science blogs. We took this initial logistic as a rough idea. We sent it to Technorati who did a preliminary ranking on it, using their ranking algorithm which basically measures the number of links to a site over the past six months, and it was also quite surprising in that, among our top science logs, they ranked in the top 3,500 of all blogs. That in itself is quite interesting and many of the scientists are surprised to find that actual pure science blogs like http://www.realclimate.org actually managed to be at the top of the blogosphere.
Anna Lacey: But what makes these blogs so good? Are they good sites or have they had some freak accident that's made them particularly popular?
Declan Butler: I think there are a couple of secrets to success that come out. Out of the top five, we have three blogs that are written by groups of scientists, and that's quite interesting. Realplanet.org., Cosmic Variance, which is a physics blog, and The Panda's Song, which is a blog of hacking creationism and intelligent design. So what we find here is that we have a lot of bloggers from a lot of different areas of expertise that allows you to actually generate better quality content and post more regularly.
Anna Lacey: And where do you think blogging's going to be going in the future? What's the next development?
Declan Butler: Well, I'm slightly biased, because I'm a blogger myself (http://www.declanbutler.info/blog/), so I feel that blogging has a very important role to play in the entire new media thing, and I think what we will see is that as certain, let's say, prejudices against blogging in the scientific community comes down with a new generation coming up, that blogging will become a much more influential and important way for scientists to actually get across their own research to their peers, but also to actually help get the research out to a wider audience.
Anna Lacey: Well, thanks Declan. That was Nature's Declan Butler talking about science blogs. But, going back now to Jo, there's a story this week about some more business-like approaches to saving the tiger. Nature 442, 12 (6 July 2006)
Jo Marchant: Yeah, that's right. It's a plan being announced this week by the Wildlife Conservation Society in New York, and they've said that they want to increase the number of tigers in their research sites by 50% in the next decade. The unusual thing about this is that you don't normally get conservation groups actually setting clear targets by which their success can be judged later on. You will generally just have these funds, but no clear black and white, this is what we pledge to do. If we don't achieve this, we will have failed. So that's sort of a new approach here.
Anna Lacey: How have these targets and this new approach been received?
Jo Marchant: Well, pretty well so far. $10 million has been pledged so far, and it seems to be going down quite well with the world of business. One of the WCS board members is Michael Kline, a New York businessman, and he's contributed some money to the plan, and he said to us, more organisations should set goals like this. As a venture capitalist, I believe in fact based judgements, based on the likelihood of success, so I think this idea that conservation societies are perhaps becoming a little bit accountable with their plans is certainly going down very well.
Anna Lacey: But it's all very well setting targets about increasing the number of tigers, but how are they actually planning to do it?
Jo Marchant: Well, they're being quite systematic about it. Basically, they've broken it down to what the tigers need. Will they need habitat? And then the second thing tigers need is prey, so they're looking at things like encouraging various actions to cut poaching, for example. So they're encouraging local agencies and Governments to do things like paying informers to expose illegal hunting, offering bounties for guns or tiger traps, and instituting bonuses for the most diligent rangers. So, it's breaking it down in steps, saying, what's the first thing we need to do? We need to reduce poaching and let's measure the success of that.
Anna Lacey: But is this actually a long term solution, because it's all very well setting targets of numbers in the short term. How are they going to make this sustainable?
Jo Marchant: The plan is over the next decade, so it's kind of not that short term, but I think the feeling is that this is just the way perhaps that conservation needs to be done, long term and it's not just a short-term thing. I think in general if people are putting money into conservation, into saving various species, they're going to want to know where that money's going, how it's being spent, what the results are, and I think that's a long-term approach, not just something for the next few years.
Chris Smith: That's Nature's Jo Marchant talking with Anna Lacey.This is Nature's Podcast from 6th July edition of Nature, with me, Chris Smith. If you would like to drop me a line with any feedback about this, or one of our previous programmes, then please email mailto:podcast@nature.com, and if you are on line, and you'd like to find a bit more about any of the items we're discussing this week, they're all available on our website at http://www.nature.com/nature.Now, would you be happy for your country to be defended by a class of nuclear weapon that's never been tested? Well, if you live in the US, then this could be about to happen because scientists are developing the RRW, reliable replacement warhead, to update the existing arsenal. But they won't be able to try it out. Here's Nature's Geoff Brumfiel. Nature 442, 18–21 (6 July 2006)
Geoff Brumfiel: What we've learned about is a new programme that only started last year in the US nuclear weapons complex and it's a programme to look for a new kind of warhead, or to design a new kind of warhead, that will be more reliable that the existing warheads.
Chris Smith: But to my knowledge, President Bush Snr banned nuclear testing in 1992, so does this mean this is literally going to come off of a physicist's sketch pad and straight into production without actually testing it.
Geoff Brumfiel: That's right. That's what they want to do. They want to build this warhead without testing it and the way they're proposing to do this is since this is going to be in theory a more robust warhead... that means it's going to be a little bigger and clunkier than the latest hi-tech designs that the designers made in the 60s and 70s. So what they actually are going to do is go back to older designs, look at some of those designs that have been tested and use them to get their computer simulations to come up with something totally new.
Chris Smith: How can you simulate a nuclear explosion? Have they got the computing power to do that?
Geoff Brumfiel: Well, you can't do it very easily. It takes a lot of code and a lot of hours of computers just running at breakneck speed. But they have gone to the point where they can do three dimensional calculations of how these bombs explode, and that's something that's very difficult to do.
Chris Smith: And one point you make here is that one of the things they're talking about doing is to substitute some of the materials, say, beryllium will be substituted by iron, because, in your words, it's less toxic. But is that really a consideration with a nuclear weapon?
Geoff Brumfiel: [laughs] Yes. That is an interesting question because, of course, the whole point of a nuclear weapon is to basically do as much damage as you possibly can, and so the idea of somehow making a more "benign nuclear weapon" will no doubt strike a few people as being kind of Dr Strangelovesque!
Chris Smith: It's almost like environmentally friendly toxic waste being stamped on a barrel of rubbish than you're going to send to some foreign country for recycling?
Geoff Brumfiel: Well, I suppose you can make that analogy. What the weapons design complex claims is that it does make a difference, because you have to build these things and if we can get rid of the dangerous non-nuclear materials, we're going to make the manufacturing process much safer and much cleaner.
Chris Smith: And do you think they will put their faith in an untested weapon, they will throw caution to the wind and say the US is defended by a new kind of weapon that we've never tested? Or, do you think they clandestinely will try it out?
Geoff Brumfiel: Well, I think it will be difficult for them for to clandestinely try it out, due largely to work conducted by the US Military in the 1960s, 70s and 80s. It's actually possible now to seismically detect virtually any underground nuclear test in the world, anywhere in the world. And this is part of why the testing moratorium has stayed in place, because if anyone violates it, we know. And so it would be very difficult for the US to conduct a clandestine test in the modern world. I think there's another question to ask and that is, if we build one of these bombs, is the US going to then decide, oh, we need to test it before we enter it into production? That is a question that the current Administration says, no way, but the current Administration are not the ones who will make that decision because by the time this bomb actually gets built, it will be 2012 or 2015, and there'll be a new President in the White House.
Chris Smith: That's Geoff Brumfiel with the US's new line in environmentally friendly nuclear weapons. Now from nuclear super powers to flower power and how it could help an astronomers to spot Earth-sized planets orbiting distant stars. The University of Colorado's Webster Cash says a flying petal shaped occulter, in other words a large disc that can cover up the star whilst the telescope looks for orbiting planets, could well be the answer. Nature 442, 51–53 (6 July 2006)
Webster Cash: We've been looking for a way to observe Earth-like planets around our neighbouring stars and it's got a lot of appeal. I wanted to do this since I was eight years old. The technical problem has been that because the parent star is 10 billion times brighter, the least little bit of imperfection in a telescope will swamp the signal from the planet and they're also very faint. So you need to go into space and have a very nice space telescope, but that's not enough.
Chris Smith: So what have you come up with?
Webster Cash: What we've come up is an external occulter. The idea is to get a space craft with an occulting disc of some kind on it that will fly into the line of sight between your space telescope and your distant star, and it turns out if you move it a large distance away, 20,000 kilometres and make it some tens of metres in diameter, you can occult the star without occulting the planet.
Chris Smith: It's not exactly rocket science, so why is it that no-one has been doing this so far?
Webster Cash: Well, it is rocket science literally, and there was one huge drawback to the occulter problem which has stopped this from happening for the last few decades, and that is the problem of diffraction. Light coming around an obscuration or through an aperture diffracts, which means it tends to scatter off to the side. About a year ago, I found a new approach. Instead of just a disc, it has to be shaped as a flower and what I've just published in this article is the formula for how you create that flower.
Chris Smith: Why has it got to be a flower shape? What's special about that particular configuration?
Webster Cash: What it does is it causes to light to diffract perpendicular to the disc and stay out of the dark shadow. See, what we're trying to do is create a shadow that's at least ten metres in diameter and very, very deep. It has to be 1012 deep. That's such a huge ratio that the least little bit of scatter around it can swamp your nearby planet.
Chris Smith: How big a planet would you be able to pick up using that particular occulting system?
Webster Cash: Well, it depends then on the size of the telescope. We have just done a lot of work adapting the occulter to work with the James Webb Space Telescope which will launch in 2013. And with that, we can see every major planet in the Solar system from a distance of 30 light years. We should be able to see even comets, discs of debris. We can see every major feature of a planetary system. It's very exciting. It's exactly the experiment you want to be able to do.
Chris Smith: And you're confident this is actually going to work? And I don't mean that in a patronising way. Obviously, this is untried technology - will it work?
Webster Cash: Of course. And that's our job as scientists, to go out and try it. So we built a heliostat about 40 metres long and we built some 1:1000 scale star shades of the petal shade, and we were able to show very deep suppression of the direct beam into the shadow. In other words, we've established that this works.
Chris Smith: Webster Cash. And that device will be deployed, all being well, with the James Webb Space Telescope which will be blasting off in 2013. Now finally this week, the sound of silence... not something that you're normally accustomed to hearing when there's an earthquake going on. But silent quakes do indeed exist. In essence, they're quakes without shakes. Or so we thought, because now Peter Cervelli and his colleagues working in Kilauea Volcano in Hawaii have found that the silent events effectively brush the seismic activity under the carpet. And they provoke aftershocks elsewhere. Nature 442, 71–74 (6 July 2006)
Peter Cervelli: In this paper which I co-authored with Paul Segall, the principle author, and Emily Desmarais, David Shelley, and Asta Miklius, we talked about earthquakes that are triggered by silent slip events which are kind of a very special type of earthquake on Kilauea Volcano in Hawaii.
Chris Smith: In what way are they silent?
Peter Cervelli: Well, originally we discovered these events with continuous GPS recorders. These instruments can establish their position on the ground to a great precision, in some cases in the order of a millimetre or two. And we detected a transient motion of many GPS receivers on the big island of Hawaii, and we were able to analyse that data and infer that a fault had slipped. And originally we thought these events were completely silent, that is, the fault slip occurred without any kind of seismicity. But now, in this recent work, we've discovered that the events themselves are silent, but they actually cause stresses within the Earth that result in normal earthquakes. We've detected these after shocks, if you will, of the slow slip or silent slip events, and the after shock locations have enabled us to get a good handle on the depth to the structure that slips during the silent event. Why is that important? Well, there are silent events that are occurring all over the world. They've been detected where people have set out continuous GPS networks, but it's been difficult to estimate the precise depth where these events have occurred. Since we now believe that the silent slip events trigger after shocks and we can locate the after shocks, we can then carefully and precisely infer the depth of the silent events.
Chris Smith: So the mechanisms that produce movements are analogous and identical to the same mechanisms that would underlie a big normal earthquake? I say normal, as in non silent?
Peter Cervelli: Well, we believe that the structure that the silent events is occurring on, the fault that they are occurring on, is also capable of producing normal earthquakes. It's a kind of puzzle how these faults can produce silent earthquakes, normal earthquakes and then also steady creep, that is slow motion of a few centimetres a year that these faults exhibit over long periods of time.
Chris Smith: So, really what's the difference between just gentle creep and one of these silent quakes?
Peter Cervelli: The gentle creep that's occurring in Kilauea is actually some of the fastest creep in the world and might be in the order of 10 centimetres a year. These silent events are occurring at rates of about ten centimetres a day and a normal earthquake is going to occur at rates approaching kilometres per second, so you have a very large range of rates of slip.
Chris Smith: Peter Cervelli from the US Geological Survey. Well, that's it for this week and thanks for listening. Next time, I shall be eavesdropping on the activity of cells in the brain's motor areas as it plans out movements. For more science news in the meantime, check out the Nature newsblog, that's available from http://www.nature.com/blogs/ and also in this week's edition of The Naked Scientists podcast, we're putting the X in sex, with a look at how at the X chromosome has evolved in the various diseases that it's linked to. We'll also be hearing about a new molecular method to find out who is eating who underneath the ground. That's The Naked Scientist Podcast which is freely available from http://www.thenakedscientists.com.Additional production this week was by Derek Thorne and Anna Lacey and I'm Chris Smith. Until next week, goodbye. AdvertisementThe Nature Podcast is sponsored by Bio-Rad, at the centre of scientific discovery for over 50 years and on the Web at http://www.www.discover.bio-rad.com.
