Nature Podcast

This is a transcript of the 06th November edition of the weekly Nature Podcast. Audio files for the current show and archive episodes can be accessed from the Nature Podcast index page (, which also contains details on how to subscribe to the Nature Podcast for FREE, and has troubleshooting top-tips. Send us your feedback to


The Nature Podcast brought to you by Bio-Rad's 1000-series thermal cycling platform. When you re-think PCR you think about how easy it can be. See how at

End Advertisement

Adam Rutherford: This week genomes get personal.

Elaine R. Mardis: The idea is not to just say, hey! We can sequence a cancer genome, you know, isn't this cool. It is cool, but at the end of the day what we want to do is have an influence on patient care.

Adam Rutherford: We find out what it means to sequence a cancer genome.

Kerri Smith: And climate change means that the natural peaks and troughs in lemming populations might be a thing of the past.

Nils Christian Stenseth: I think the huge lemming peaks will be gone for ever.

Kerri Smith: This is the Nature Podcast. I'm Kerri Smith.

Adam Rutherford: And I'm Adam Rutherford. Just seven years after the human genome we are truly entering the era of personal genomics. Earlier in 2008, we published the first individual genome. It was DNA pioneer Jim Watson's in fact. The technique reduced the time and money taken from years and millions to just weeks and hundreds. In this week's Nature, genomics has been taken to the next level with the publication of three personal genomes. We have the sequences of an African and an Asian individual, but also the genome of a single cancer patient. A team from Washington University in Seattle identified 10 mutations in the cancerous tissue of a patient with acute myeloid leukaemia or AML, in order to determine how cancer progresses at the most basic genetic level. I spoke to one of the authors, Elaine Mardis and started by asking her what it means to have 10 distinct DNA mutations in one cancer. Nature 456, 66–72 (6 November 2008)

Elaine R. Mardis: The basic premise is that the mutations that transition in normal cell population to a cancerous cell population are the types of mutations that one incurs as time goes on. Cancer is often described typically as a disease of older people and relive what that description invokes is that throughout your life because of mistakes in DNA replication that aren't being caught by the machinery that catches such things, you cumulate mutations. So for this patient, it was a combination of those acquired mutations that led to the tumour as well as we think from her family history, a general tendency to develop cancer, specifically blood born cancers.

Adam Rutherford: Okay, so in this case this individual and that disease the acute myeloid leukaemia, they have acquired 10 specific mutations, now from you paper you knew a couple of those mutations already, but the rest were novel. How different is it going to be from this individual to another who has the same condition?

Elaine R. Mardis: That's a great question. I think our early indications from looking at the mutations that we found in this individual against a panel of other tumour DNAs acquired from other AML patients was that there were not similarities within those specific genes. So on the one hand that's a little surprising, perhaps disappointing but when we look at our other studies of cancer where we are not sequencing the entire genome but rather focusing in on specific genes, what we find is that even for a limited number of genes the number of genes that are similarly mutated across patients for given tumour type is generally fairly low. Rather what seems to be impacted most by the mutations are the pathways, the cellular pathways that the genes that are being mutated or participating in or more properly stated that their protein products are participating in.

Adam Rutherford: Okay, so how far are we from developing diagnostics based on personal genomics of cancers?

Elaine R. Mardis: You know, I think one of the dangers in all of these sorts of things is that it comes across as being sort of the answer and it's really just sort of the beginning piece that we have of information. The reason that we try to publish this is so that people were aware that this approach was available, AML patients now are treated basically the same as they were 25 years ago. There really aren't any differences in the treatments and we don't really understand what, you know, how one patient will do versus another patient when they initially present with the disease, so everyone is given the same course of treatment. Ultimately, we would like to be able to stratify that.

Adam Rutherford: And do you think that this level of personal genomics is the first step into that really precise individual profiles of cancer patients, not just AML but across the board.

Elaine R. Mardis: I absolutely do think that that is the case and it begins to address a really interesting question along the lines of personalized genomics. In other words, at the end of these additional genomes that we sequence, are we going to emerge with paradigm that clearly makes it important for each patient to have their tumour genome sequenced, so that they get the best, you know, most personalized care or as we sequence more and more will there be commonalities that emerge that allow us to focus our attention on sequencing only specific genes or sets of genes and then diving off from those mutational spectra into the exact treatment for that, you know for that specific series of gene mutations. So in other words, it won't be as highly personalized as one at a time genome sequencing would require, but rather will allow us to immediately stratify patients into one or more subgroups that we can then use tailor therapies, you know, in a broader sense. The idea is not to just say, hey! We can sequence a cancer genome, you know, isn't this cool. It is cool, but at the end of the day what we want to do is have an influence on patient care.

Adam Rutherford: Elaine Mardis from the University of Washington. Those other personal genome papers one Asian individual and one African as well as analysis and features are freely available at

Kerri Smith: Well, as it is becoming clear there is a whole genetics jamboree going on in the magazine this week. The package includes a feature from Editor Brendan Maher about the quest for the missing heritability. No, not the next Harry Potter movie, but almost as mysterious. Luckily Brendan is on hand to explain. He is on the line from Philadelphia. Hi, Brendan. Published online 5 November 2008 Nature 456, 18–21 (2008)

Brendan Maher: Hi.

Kerri Smith: Tell us first of all, what do geneticists mean by heritability?

Brendan Maher: Heritability is really a measure of how much variability in the population is transmitted by genes as opposed to picked up from environment.

Kerri Smith: And what about it is missing then in your feature?

Brendan Maher: So, a really good example is height. Now this summer three groups did a huge genome wide association studies and what they found was about 40 genes that account for differences in height between people, but the catch was all these genes only account for about 6 or 7 cm difference between people, so researchers were looking and saying where are the rest of these genes that account for the vast variability in height that seems to be in people.

Kerri Smith: And they thought previously then from twin studies and from other ways of analyzing these traits that it was up to sort of 90% heritable.

Brendan Maher: That's right. They only accounted for may be 5% of that.

Kerri Smith: Mmmm... confusing!!! So, for your feature then you spoke to several geneticists and asked them where they think this missing heritability could be found and some of them said it's probably right under our noses and have we not been looking hard enough.

Brendan Maher: Not exactly, it's just that the way these genome-wide association studies pick up genes and pick up loci that are on the genome and that doesn't necessarily point to what the underlying heritability conferred by a specific gene might be. So for example to me, two people may look exactly the same at a locus based on these genome-wide association studies, but if you really go down and start sequencing their gene specifically, you might find minor changes that have really drastic effects in height and that all kind of gets washed out by the genome-wide association studies. But there is another place where heritability might be hiding and that's genes that we just don't know anything about right now and that will probably take much more sequencing, at least according to some people across just broad swathes of the genomes have defined variants that have never really turned up from genome-wide association studies.

Kerri Smith: So, some of it is that we just haven't come across certain genes that might be involved in particular traits, but what if it isn't actually about finding anymore genes per se, there must be other ways in which traits are passed on?

Brendan Maher: Right. So, well, one thing people are looking at is genetic networks. It is no surprise that genes don't always act alone. One gene doesn't necessarily mean one protein or one phenotype, but genes can actually work in concert, sometimes in very complex ways: epistasis, for one thing, can allow one gene to completely mask the effect of another or it can mean that two genes that have a very small effect on their own could have a much larger effect when they appear together and we don't really have very clear models of how this might be working in human genetics. Animal models kind of point us in the right directions in some ways, but gene-gene interaction is a really important factor and that doesn't even take into account gene-environment interaction because the environment can really impact the way genes are working in profound ways.

Kerri Smith: You gave the example of heights in the article as a trait that isn't very well explained in terms of heritability, so I wondered if you could just give us a couple more examples of other things where there is a massive gap between what we know about genes and what we think the heritability should be?

Brendan Maher: Right, the neurological diseases have been certainly bugaboo. The genome-wide association studies haven't really picked up much of anything.

Kerri Smith: And one geneticist who works on the genetics of autism, which is one of these diseases, I gather, whose heritability isn't very well explained is Dan Geschwind of UCLA, so he had this to say about the effects of genetics in disease research.

Daniel Geschwind: Genetics is changing our definition of diseases in general and breaking down some of the current notions of disease boundaries and to certainly in a disease that's as complex and in some ways ill defined as autism and is broadly defined, the genetics is likely to change the way we look at it. In other words, you can now identify a newer parent, a particular kind of mutation, your child then becomes that kind of disease that has certain prognostic implications that we don't understand yet, but may also have treatment implications as well.

Kerri Smith: Brendan, do you think our ways of diagnosing complex conditions like autism is causing us to loose heritability somewhere along the way?

Brendan Maher: Well, some people say that by lumping all diseases together into a single diagnosis, schizophrenia is a big one obviously and then autism, kind of undermines the ability of genetics to really tell us much about them. Other people disagree with that.

Kerri Smith: So let's widen the picture a bit and here about what else is going on in the genetics special in the magazine and indeed it is possible now for anyone to have their own genome sequenced at the snip of under 500 dollars. When I spoke to Dan Geschwind, he had some concerns about personal genomics.

Daniel Geschwind: The communication of what genetic risk means even by a trained professional to very educated people is extremely complex and therefore expecting people to understand the complexity of their genome and its medical implications without significant interaction with professionals doesn't make much sense to me.

Kerri Smith: Now Brendan, there is a commentary in Nature by a team led by Barbara Prainsack at Kings College, London that addresses just this issue of personal genomics, did the authors agree with Geschwind?

Brendan Maher: Well, really their argument is that we need a little more time and little more effort in looking at what exactly these sorts of companies are doing to people and how people interact with them. It's not necessarily clear that any sort of harm has come from someone having their genotype to add a couple of hundred different snips. The information that they provide, it's however pretty weak predicted power more like fortune telling in some cases than really predicting disease. So they are urging for a little more caution rather than trying to put out precautionary regulation without knowing all effects.

Kerri Smith: Okay, thanks for talking to us Brendan.

Brendan Maher: Thanks Kerri.

Kerri Smith: And that's before you can find more of the special genetics coverage online at

End Jingle

Adam Rutherford: Now here's Geoff Brumfiel with a totally different matter.

Geoff Brumfiel: Dark matter is a mystery material that makes up about 85% of the matter in the universe. Our own Milky Way is full of this stuff, but astronomers haven't yet been able to see up with their telescopes because it's, well dark. As far as we know, it only interacts with visible matter via the force of gravity, but a new computer model suggests another way to search for dark matter. It predicts that the dark stuff might periodically self-destruct giving off bursts of powerful gamma rays that would be picked up by space-based telescopes. I spoke to Volker Springel of the Max Planck Institute for Astrophysics about where to look for something you can't see. Nature 456, 73–76 (6 November 2008)

Volker Springel: I mean, it's basically a hypothesis of course, the existence of dark matter is a hypothesis, so we assume that it's an elementary particle with certain properties and that the primary important property is that this particle, this dark matter particle does only seal gravity, that's the only physical interaction it participates and perhaps this very very few exceptions and actually in our paper, we studied one of these exceptions that if the dark matter particle is of special type, it might self-destruct occasionally and then you would see it.

Geoff Brumfiel: So how would you see it then, what would it look like when it self-destructed?

Volker Springel: Yeah, so it's a very heavy particle and that would mean that if it self-destructs; it generates very hard radiations, gamma rays. These are more energetic even then x-rays. These gamma rays would then be emitted where these particles find each other and destroy each other and would emit photons, these photons are not very many, but you would basically need a gamma-ray detector telescope and you would then see on the sky, a faint glow of gamma rays from occasional annihilation from dark matter particle and the question then is where would this emission be seen in the galaxy or from other sources and where you should look for it.

Geoff Brumfiel: That's what you've done in this paper, I understand. You've actually performed some supercomputer calculations to try and paint a picture of what this gamma-ray glow might look like.

Volker Springel: Yeah, that's true. Right, we simulated the formation of the Milky Way Galaxy and that allowed us to come up with a very detailed picture of this, where the dark matter is in our galaxy and how it is distributed, because the special thing about dark matter is also that it is very clumpy and lumpy. It accumulates in tiny clumps, millions of them, in fact billions actually that over time grow bigger and bigger and merge and make up then what we call the halo of the Milky Way, this is kind of a big ball of dark matter in which our galaxy sits. The interesting physical question now is will these little dense clumps, will they produce a lot of gamma rays and the big question in astrophysics is for us now that NASA's newest observatory is able to look for these gamma rays, whether it should look for the emission coming from these small little lumps or whether it should look at another place, for example at the galactic centre based on the emission from the main halo of the Milky Way.

Geoff Brumfiel: And so what did you conclude?

Volker Springel: So, our conclusion was basically that it's better in fact to look for the emission from the main halo of the Milky Way, not by doing it by directly looking into the galactic centre. There you would risk being confused by other sources. There is a black hole for example in the galactic centre which emits hard radiation and other strange stars, quasars, so and so, but we still find that it's the best strategy to look about 10 to 30 degrees away from our galactic centre and that is a conclusion that's at odds with some claims and estimates that have been made earlier.

Geoff Brumfiel: You know, I have to ask I mean, we are talking about a particle that has never been directly seen and self-destructing in a way that it may or may not do and guessing with a computer model, why should we believe any of this, I mean are you just taking jots on the dark.

Volker Springel: That's certainly a good question. Well, the dark sector in the universe is of course amidst of we have this strange universe that allegedly we say consists mostly of dark matter. The same computer model that we used here they are able to accurately predict the distribution of galaxies in space, the dynamics of galaxy clusters, the properties of galaxy, their size distribution, their clustering pattern and so on. So it's an enormously successful description of what do we really see, so these computer simulations have been tested. Now we have only noticed once this emission is discovered or not, whether this is actually true.

Adam Rutherford: Volker Springel talking to Geoff.

Kerri Smith: Next up, climate change could be getting to the most iconic of rodents, the lemming. Reporter Katherine Sanderson has the story.

Katherine Sanderson: Lemmings are the little rodents famous for allegedly committing mass suicide and also for starring in their very own computer game, but these arctic creatures are now under close scrutiny by scientists looking at the effects of our planet's rapidly changing climate. I spoke to Nils Christian Stenseth from the University of Oslo. He was being pondering the case of the missing lemmings. Do lemmings really leap off cliffs in acts of mass suicide? Nature 456, 93–97 (6 November 2008)

Nils Christian Stenseth: Well, that is unfortunately not true. They do not march to the sea and commit suicide but they do increase in high numbers every once in a while and they are paused every 3 to 4 years and when there are lemmings all over and there is a hill side the habitants, even though walk around to sort of move downward and downward is really down to the sea.

Katherine Sanderson: Okay, so the bit about suicide isn't true, but as you mentioned lemmings do seem to just appear in huge numbers and then disappear again or at least they used to, can you sort of tell us what has happened over the past 30 years to lemmings cycles?

Nils Christian Stenseth: Well, up until mid 90s we had lemming peaks every 3 to 4 years in the mountains of Norway and really in the whole Scandinavia. It was a very regular phenomenon. People knew about this and they sort of look forward to it and was very excited and kids were afraid. They are not very aggressive but they are screaming. Then in 1994, the lemming's cycles and for that matter other small rodent cycles started disappearing in Scandinavia and people were speculating that it might have a link to the climate, because climate was changing as we know at the same time, but no one really had any mechanism and suggestion for how this was brought about.

Katherine Sanderson: So to put it in context a little, what kind of effect does the disappearance of this cycle have on the ecosystem in Scandinavia?

Nils Christian Stenseth: It has a huge effect on the ecosystem, because when there is high numbers of lemmings, foxes, and birds of prey they have very good conditions, so they can feed lots of young. Then the next year when the lemmings are gone, Ptarmigan and other species they will then be preyed upon by birds of prey and foxes. So the whole dynamic in this high mountain system will disappear and as a matter of fact the ptarmigan and other species might actually suffer greatly from this.

Katherine Sanderson: And tell us how you linked these changing lemming cycles or disappearing lemming cycles to climate change?

Nils Christian Stenseth: Well, I have been studying lemming for eight years actually since 1970s while I did field work and I have been continuing analyzing data on lemmings and other small rodents and I also was part of taking snow measurements. These data were not continued being collected as we started up as a research program, however, the students of course at the mountain hill station, they were taking samples of the same kind and these data were sort of stored on the shelf, and well we have sort of forgotten about this and then my student, the lead author Kyrre Kausrud, he was taken this course and he realized that may be we should include this data in the analysis and basically that become the start of the paper that is coming out in Nature now.

Katherine Sanderson: And tell us what you saw in the changes in the snow? I mean, you talk about wrong snow, I think.

Nils Christian Stenseth: A change in the snow is that it is not the amount of snow that has changed but it is the kind of snow that has changed. Earlier on, the snow was soft so that a sub median space of 5 cm or so, between the ground and the snow could develop and within that space, the lemmings were reproducing and surviving very well. So they could build up a population during part of the winter and then continue to develop the population or increase the population in the subsequent summer. When climate started changing in the mid 90s, it was more wetter snow coming so that it was not so easier to develop this space between the ground and the snow so that was really what we saw in the data and they could really explain the dynamics of the lemmings, the changing dynamics using this data.

Katherine Sanderson: And what happens now for lemmings, what do you predict is their fate in the coming years?

Nils Christian Stenseth: Well, one should always be very careful about predicting. However, my feeling is that the lemming cycles might be a phenomenon of the past because there is no question that the climate will continue changing in the direction it is doing in the various parts of the world and I think, the huge lemming peaks will be gone for ever.

Kerri Smith: Nils Christian Stenseth talking to Katherine Sanderson.

Adam Rutherford: Okay, before we wrap it up for another week, News Editor Mark Peplow has joined us in the pod. We are recording on Wednesday morning, the results are now in and Barack Obama will be the next US President. Now what is that going to mean for science, Mark.

Mark Peplow: It's potentially very exciting for science. You remember that Nature actually endorsed Barack Obama's presidential candidate, it is the first time that Nature has ever endorsed one, so I can officially say Hurray! that he has won. He promised just a whole host of changes, new investments in science and technology over the course of his campaign and he had a very wide coterie of science advisors feeding in to the campaign and putting very well thought out policy proposals into his planning.

Adam Rutherford: And the size of the victory is that going to allow the Democrats and Obama to push through bills which are going to be radical for science?

Mark Peplow: Well, looking at what the New York Times latest results at the moment, you know, there is a fairly convincing majority in the House of Representatives. Obama needed 270 Electoral College votes to actually win and has 338 at the moment at that time that we were recording this. The senate they don't quite have that filibuster-proof majority that would allow the democrats to push through anything they like, but I think it is very likely that they do have the strength that they will be able to push through democratic priorities such as climate change legislation and if Obama actually carries through on that when he gets into office in January next year.

Adam Rutherford: Okay, so you have mentioned climate change, what are the other big issues for scientists that the democrats will try and push through?

Mark Peplow: Well, energy and environment is a huge issue and you know in his acceptance speech last night, Obama cited a planet in parole among the leading challenges that his presidency is going to face and he did promise a 150 billion dollar push in new energy research and that's something which is obviously going to have a massive impact on science and if that actually goes through as promised.

Adam Rutherford: Okay, we will look forward to seeing some of those things implemented over the next four years. Now what else is in the news, I understand that the Phoenix is fading away.

Mark Peplow: Phoenix is fading away, Phoenix is a Mars Lander and NASA landed on Mars earlier this year and it landed in one of the polar regions of Mars, specifically to try and investigate the water ice there. Because it is solar powered and the Martian winter is approaching in that part of the planet, it now is getting pretty much to the end of its life. Over the last week its communication with Earth has been very intermittent and it shut itself down a couple of times into like a hibernation mode and it's on its last legs. One of the people that we spoke to on the mission seems that you know it's like an aging parent in a nursing home. You know the end is coming.

Adam Rutherford: And has it fulfilled its mission, has it generated the data that the scientists wanted to see from Phoenix.

Mark Peplow: Well, the story is being quite mixed actually. There have been some frustrations with some of the instruments. At the start of the mission, they did have some success, they clogged their way through the top layer of few centimetres of soil to actually find water ice there, now water ice have been observed at the poles from orbiting instruments before, but this is sort of ground truth and it means a lot to the sort of planetary geologists who are actually running this machine.

Adam Rutherford: But it seems that the main experiment that they were hoping for didn't actually work for Phoenix.

Mark Peplow: Yeah, I mean having some real problems and it had a suite of eight ovens that would take in soil and then bake them and sniff the emitted gases for organic compounds. The idea was that you can see what organic matter is in the soil. You can even potentially start to look at isotope ratios to see what sort of processes those compounds have been through, but the truth is that they wasted nearly half of the mission trying to get those ovens to work, the soil was very sticky, they can't get the soil in, they had to shake the ovens to try and sift it all in and that caused some electrical problems. It was kind of a nightmare and in the end they never actually got to test the ice for the isotopic ratios that could have said something about its age. So that's been a kind of disappointment really that that didn't go very smoothly, but they do have a huge swathe of data now and which because the scientists have been so wrapped up in actually running the machine, now they are just digging into that data and over the coming months we will be seeing conclusions from that.

Adam Rutherford: Okay and if you are interested Phoenix has been twittering, that's the instant messaging service from the surface of Mars, so you can see that death as it happens. Now Mark, nanotechnology has a practical use in speaker technology.

Mark Peplow: Yeah, that's right. Nanotech is cropping up now in all sorts of different products, but this is a really really nice story. I liked this and Chinese researchers have found that if you take a sheet of carbon nanotubes and each of them just 10 nanometres across, you can actually turn them into a loud speaker. Basically, this sheet, very very thin is transparent. When they apply an electric current which is alternating an audio frequency, it actually produces a sound as loud as commercial speakers.

Adam Rutherford: And we have got a recording of that sound as played through the new type of speakers.(Playing Beijing 2008 Olympic theme)

Adam Rutherford: Okay, that was the Beijing 2008 Olympic theme. Now, that sounds a little bit tiny to me.

Mark Peplow: Well, that's fair enough. I mean that, you know, they are not high quality board speakers these are like tiny thin sheets that you can potentially wrap on to clothes and they actually stretch to on over the screen of an iPod and plugged into the iPod so that the sound is actually amplified out of that and you can see stuff through the screen. So, I mean you can imagine a whole host of applications to this sort of thing. It's actually working not because the sheet itself is vibrating to produce the sound, but it's actually as the current flows through the nanotubes and they alternately heat and cool down which heats and expands the air near them and it's that heating and cooling that causes the sound waves. We talked to a guy called Howard Schmidt at Wright University who works in nanotechnology and he said it's so wonderfully simple. It brings up the strong wave of 'Duh, why didn't I think of that!'.

Adam Rutherford: Fantastic Mark, all of those stories and much more available on That's it for this week's show.

Kerri Smith: Next week, it's the Society for Neuroscience Mega-conference, so we will be bringing you some brainy bits and bobs from the journal. I'm Kerri Smith.

Adam Rutherford: And I'm Adam Rutherford.


The Nature Podcast is sponsored by Bio-Rad, at the centre of scientific discovery for over 50 years. On the web at

End Advertisement