[Jingle]
Interviewer: Shamini Bundell
Welcome back to the Nature Podcast. This week we’re finding out how to reduce cancer risk for people with red hair and how machine learning is speeding up our analysis of the universe’s gravitational distortions.
Interviewer: Adam Levy
Plus, an update on a big story from a few weeks ago. This is the Nature Podcastfor September the 7th2017. I’m Adam Levy
Interviewer: Shamini Bundell
And I’m Shamini Bundell
[Jingle]
Interviewer: Shamini Bundell
First up this week, Adam’s been looking into a problem that has been plaguing researchers around the world.
Interviewer: Adam Levy
Kelly Cobey is the publications officer in Ottawa Hospital in Canada. She deals with all sorts of issues researchers might have but one researcher stands out. He’d received an invitation to submit a paper to a journal, so he went ahead and submitted a paper.
Interviewee: Kelly Cobey
And shortly thereafter his manuscript was accepted. There was no peer review and he was asked to pay an article processing charge for that. So, at this point alarm bells went off. He knew something wasn’t right. This is not the typical process.
Interviewer: Adam Levy
He was shaken up and quickly tried to retract his manuscript, but to his astonishment the journal told him he’d have to pay a retraction fee to avoid the piece being published.
Interviewee: Kelly Cobey
The journal eventually actually went forward and did publish that paper without his consent. It was a very frustrating situation for this author.
Interviewer: Adam Levy
This researcher had fallen prey to a predatory journal. And he’s far from alone. There’s growing concern about these kinds of unscrupulous publications. But what actually is a predatory journal?
Interviewee: David Moher
Predatory journals are perhaps explained by their behaviour, or I would say, perhaps by their misbehaviour.
Interviewer: Adam Levy
This is David Moher who works on publication science also at Ottawa hospital. David tells me that predatory journals can misbehave in a plethora of ways. They may claim to do peer review while really having no such process. Members of the journal’s editorial board may not even have heard of the journal. In one particularly extreme case…
Interviewee: David Moher
There was an editor in chief or a journal who, unfortunately, the gentleman was dead for two years.
Interviewer: Adam Levy
Although predatory journals are defined by dodgy practices, what defines the research they publish has remained unclear. And so David and collaborators have laboriously studied almost 2,000 biomedical articles published in predatory journals. And these papers reveal some unexpected patterns.
Interviewee: David Moher
I think our biggest surprise is that these journals are populated by research coming from all nations. There was this belief that perhaps most of what was appearing in these journals was coming from low-middle income countries or countries who couldn’t afford to publish elsewhere.
Interviewer: Adam Levy
And the surprises didn’t end there. There were publications from authors at top institutions with prestigious funding sources. In fact, in David’s sample, the US National Institutes of Health was the most commonly named funding source. But while the research may come from well-respected institutions, the papers themselves were often of a strikingly low standard. Papers often omitted to mention how samples were randomised, for example. And this wasn’t the only thing that they missed out.
Interviewee: David Moher
A fair number of reports who should have been very explicit about having consent and ethics approval were missing and that speaks to the broader issue of editorial oversight in these entities.
Interviewer: Adam Levy
To David, these insights reveal the harm caused by these publications.
Interviewee: David Moher
The bottom line is that for funders this is extremely wasteful. It’s wasteful in terms of human resources; people participating, and it’s a waste of resources.
Interviewer: Adam Levy
So what can be done to reduce this waste? David is key to stress that we don’t yet have data on why academics submit to these journals in the first place but Kelly tells me that the man who realised too late that he had submitted to a predatory journal is far from unique. Kelly comes across many researchers who are unaware of the warning signs.
Interviewee: Kelly Cobey
When I cover predatory journals, giving seminars, inevitably I’ll get emails after the lecture or folks will come up to me after the lecture starting to wonder about the integrity of where they submitted and that’s been a reoccurring theme after giving outreach.
Interviewer: Adam Levy
So, one way of combatting predatory journals is simply for more academics to know how to spot them. One paper in BMC Medicine identifies thirteen characteristics of predatory journals but Kelly says that one of the most important things a researcher can do is be wary of their inbox.
Interviewee: Kelly Cobey
It’s quite rare to get a solicitation from a journal that you’ve never heard of, so that would be a red flag for me.
Interviewer: Adam Levy
Raising awareness of red flags like this would help. But David feels like there are many other steps that could and should be taken to fight back against predatory journals.
Interviewee: David Moher
Academic institutions and funders need to come together and develop policies and procedures and put them in place to really work at making sure that people do not submit to these entities. And so if we have a combination of more research, training and policies, I think if we take that as a poly-intervention, then I think we will really stop predatory journals in their movement forward.
Interviewer: Adam Levy
That was David Moher and Kelly Cobey. David’s written a Comment piece and Kelly’s written a World View on predatory journals and you can find both at nature.com/news. To help spot predatory journals yourself, check out the paper in BMC Medicine. That’s Shamseer et al. and published in 2017.
[Jingle]
Interviewer: Shamini Bundell
The most common human hair colour is black, then there’s a spectrum all the way through brown to light blonde. But around one in a hundred people have naturally red hair. It’s associated with fair skin, freckles and is more common in people whose ancestors came from Northern or Western Europe. It’s also associated with a particular medical problem, a problem that Rutao Cui wanted to solve.
Interviewee: Rutao Cui
My study is trying to understand and find a way to prevent a skin cancer, a most malignant form of skin cancer – melanoma – in red haired individuals. Red haired individuals are frequently diagnosed with melanoma.
Interviewer: Shamini Bundell
Your chances of developing melanoma, the most dangerous type of skin cancer are significantly increased if you have red hair. That’s why Rutao and his colleagues wanted to study the mechanisms behind this link. Their research is being published in Naturethis week along with a News and Views article and I spoke to one of the News and Views authors, Liz Patton of the MRC Human Genetics Unit at Edinburgh about the research. I asked her why red haired people are at a higher risk of developing melanoma.
Interviewee: Liz Patton
So it’s because they have a mutation in this gene called MC1R and this is a receptor that responds to signals following UV damage. So the UV from the sun goes on to our skin and that activates the MC1R receptor and one of the things is does is it stimulates pigmentation so that’s why we get a tan and so people that have an MC1R mutation, rather than making a dark brown pigment to protect the skin and have a tan, they form a red pigment and they freckle and they don’t tan very well.
Interviewer: Shamini Bundell
So this is why my red-haired friend complains that he just gets sunburnt?
Interviewee: Liz Patton
Exactly, so they don’t have the potential to tan and what’s quite interesting about it is that even having one copy of the MC1R variant, or mutation, can put you at greater risk of melanoma.
Interviewer: Shamini Bundell
Right, so my red haired grandmother could make me more at risk of developing melanoma?
Interviewee: Liz Patton
Yes, exactly.
Interviewer: Shamini Bundell
Darn, none of us are safe. And so the MC1R is linked specifically to melanin, that brown tint that you get when you tan?
Interviewee: Liz Patton
Yes.
Interviewer: Shamini Bundell
And this team wanted to figure out exactly how the MC1R protein is affecting the melanin production in your skin and affecting melanoma formation. And what did they discover about how it actually works?
Interviewee: Liz Patton
So what they discovered was that at the end of this MC1R receptor is a site where the receptor can be palmitoylated, so it adds a lipid. And what they found was that people with red hair and poor ability to tan and increased freckling, they found that they have lower levels of palmitoylation.
Interviewer: Shamini Bundell
So this is the word that I’ve being trying to practise pronouncing: palmitoylation.
Interviewee: Liz Patton
Yes.
Interviewer: Shamini Bundell
And it’s a modification of the protein and it’s what’s missing in people with this gene variant: that there’s not enough palmitoylation happening and that’s what seems to be causing all the defects, so the skin not tanning properly, not providing proper protection from UV rays. Now that they’ve figured out that that’s how it works, can we use this understanding to actually prevent melanoma?
Interviewee: Liz Patton
Yes and for me this is what’s really exciting about their paper; they can use small molecules or drug-like compounds to increase the palmitoylation. And then rescue the defects in the MC1R variant. So to me that’s really exciting because that suggests that you could use a drug to help prevent some of the phenotypes associated with this mutation.
Interviewer: Shamini Bundell
And they actually tested this on mice and they got mice with the red haired gene?
Interviewee: Liz Patton
That’s right, they engineered a mouse that had a variant that is found in people and then this mouse also has sort of the red hair phenotype. What they found is that they increased palmitoylation, now the MC1R worked more like a normal variant of MC1R and had reduced risk of melanoma… reduced incidents of melanoma.
Interviewer: Shamini Bundell
So there’s a drug that works in mice, essentially, to reduce this high risk of melanoma formation.
Interviewee: Liz Patton
Yes
Interviewer: Shamini Bundell
But this is an intervention that’s preventative so obviously the dream would be, right; can we stop people getting skin cancer? Is that a practical approach?
Interviewee: Liz Patton
I mean, that’s a fascinating question. These are some of the questions that we’re now entering into in our personalised medicine genomic era. You know, should we be thinking about drugs to treat otherwise healthy people to prevent a disease that they don’t have yet? And I think that’s a really important question so I think probably it’s unlikely that you would use this as a preventative treatment on a large population scale. One case is, you could imagine if someone was badly burned, that you would want to stimulate the repair of the damage in the DNA and so this drug or other strategies that target the similar pathway might be appropriate in that context.
Interviewer: Shamini Bundell
You mention personalised medicine: so if I were to have my genome analysed and even if I don’t have red hair were to find out I’m an increased risk, I personally might want this preventative treatment?
Interviewee: Liz Patton
I guess my personal feeling is that we know there are sunburn prevention strategies, you know, wearing clothing and sunscreen and those are really known ways to prevent, to reduce your risk of melanoma.
Interviewer: Shamini Bundell
So that’s probably still a safer bet?
Interviewee: Liz Patton
Yes, absolutely.
Interviewer: Shamini Bundell
Don’t get sunburn basically.
Interviewee: Liz Patton
Yes.
Interviewer: Shamini Bundell
You and your co-author mention at the very end of your News and Views, are there sort of natural products that could have the same effects?
Interviewee: Liz Patton
Yeah, right. I mean, when I was young people used to put on coconut oil and this kind of thing when they were tanning and I always thought that was to make their skin kind of not dry out but actually people claim that it can give a better tan and there has been an publication about palmitic-acid ester in some natural products that can stimulate melanin in melanocytes. So it makes you wonder if there are some of these natural products out there that are stimulating the tanning response.
Interviewer: Shamini Bundell
Further research, that’s the next stage then, whether we should all be slathering ourselves in coconut oil? But as well as, not instead of, sun cream right?
Interviewee: Liz Patton
Yes.
Interviewer: Shamini Bundell
This is turning into a public service announcement about telling people to wear sun screen.
Interviewee: Liz Patton
Yes, wear sun screen, wear clothing.
Interviewer: Shamini Bundell
That was Liz Patton from the University of Edinburgh with her top tips for avoiding sunburn, especially likely if you have lighter coloured skin. You also heard from study author Rutao Cui who’s based at the Boston University School of Medicine. You can read both the paper and the News and Views article at nature.com/nature.
Interviewer: Adam Levy
If you were paying attention to the science news a few weeks ago, or even if you weren’t, you probably heard about a new study on gene editing human embryos. Well, since then, some doubts have been raised about this work. We’ll take a look at the latest on this story in the News Chat at the end of the show.
Interviewer: Shamini Bundell
Now though, it’s time for two new pieces of research. It’s this week’s Research Highlights with Charlotte Stoddart.
[Jingle]
Interviewer: Charlotte Stoddart
A normally asexual organism has been thrown into a mating frenzy by a bacterial protein. Choanoflagellates are the single celled organisms most closely related to animals. They usually multiply by dividing themselves so don’t need a sexual partner. However, when a marine bacterium was let loose in a dish of Choanoflagellates, they got a bit frisky, clustered together, and reproduced sexually. It turned out the bacterium was secreting a protein that acted like an aphrodisiac on the normally celibate flagellates. Naturally, the researchers named the protein Eros after the Greek God of sex appeal. More on this raunchy research in Cell.
[Jingle]
Interviewer: Charlotte Stoddart
The lure of moving to Canada is spreading to some rather unwelcome insects. Southern pine beetles are destructive denizens of America’s Deep South. They burrow into the bark of fir trees, sometimes proving fatal to entire forests. Their happy chomping ground is usually restricted by an invisible wall of cold weather to the north. But a study of recent beetle expansion in New Jersey revealed that beetles scuttle into new areas when winter temperatures stay above minus ten degrees Celsius for ten years. If the climate warms as models predict, these pests could be laying waste to woodlands in Canada within the next few decades. More on the details in Nature Climate Change.
Interviewer: Adam Levy
When you picture a lens, you might think of a camera lens, a microscope lens, or maybe even an eye. But you probably don’t think of a galaxy trillions of miles away from Earth. But galaxies and other massive objects can act as gravitational lenses, distorting the light from galaxies elsewhere in the universe. And now scientists have developed a new method of analysing these light distortions. Reporter Anand Jagatia spoke to astrophysicist Yashar Hezaveh from Stamford University. Anand started by asking him how gravitational lenses actually work.
Interviewee: Yashar Hezaveh
If you ever look at an image of something with a magnifying glass the image looks distorted and the reason that the image looks different and distorted is that the light rays coming from that background object are getting bent by the glass because of this refraction. A similar phenomenon happens when we look in the skies but because of gravity. So we have two galaxies – one galaxy in the middle, we call it the lens galaxy just like a magnifying glass – and a second galaxy that could be much further away. And so as the light rays of the background galaxy pass near the foreground galaxy, the lens galaxy, they get bent because of gravity. And so as a result here on the earth when we look at them we see this distorted image of the background galaxy and sometimes we even see two, three, four different images of the same background galaxy. It’s a little bit like looking in a fun house mirror.
Interviewer: Anand Jagatia
So can you describe what these galaxies actually look like once they’ve been distorted in this way?
Interviewee: Yashar Hezaveh
It’s really fascinating. You see this middle galaxy that looks like a normal galaxy – it’s elliptical, it’s spiral, something like that – and there’s this crazy ring or multiple images around it. In fact at home if you get a wine glass and look at an image of a candle through the foot of the wine glass, they produce a very similar kind of effect. You can see like a ring shape and sometimes multiple half arcs around the stem of the wine glass. It’s a very similar effect. Gravity is actually bending light and redirecting it to come to us.
Interviewer: Anand Jagatia
Is gravitational lensing a problem then if we’re trying to look at objects in the sky? I mean, if they’re getting distorted by other objects is that why it’s a problem because we can’t really get a good sense of what their structure is?
Interviewee: Yashar Hezaveh
It could be a problem but mostly it’s a great tool. We actually use gravitational lenses to look at galaxies that would otherwise be really difficult to detect in our telescopes, to observe. It magnifies them so we can look at them with much higher resolution and much sharper images. But as you mention it also distorts this image so we need to find a way to correct for this distortion. Again it’s like looking at yourself in a mirror which is distorted. Maybe the mirror makes us look a little bit skinnier or a little bit chubbier and so we need to find out exactly what part of this is my own image and what part of it is the distortion caused by the mirror.
Interviewer: Anand Jagatia
So how do we go about doing that then? If we stick with the funhouse mirrors analogy, how do you go about trying to make sense of that image?
Interviewee: Yashar Hezaveh
So we say I will come up with a model for what I am, what I look like, and I will come up with a model for how the mirror is and now I need to simulate it. So I put my own image through a distortion by the mirror and produce a distorted image. Maybe I look a little bit taller. And then I compared that image to the real data that I have. And I say based on statistics and noise and things like that, how likely is it that the true answer is what I basically calculated in the first place? So we need to go through probably thousands, tens of thousands of different answers for that in our simulations and for each one of them simulate the whole thing, compare it to the data and then move on to the next one. But next time we have another problem, we need to redo the whole thing again.
Interviewer: Anand Jagatia
So I guess your approach in this paper is trying to get around some of these difficulties and you were using deep learning, right? So you were training neural networks to analyse these pictures.
Interviewee: Yashar Hezaveh
So what we used in this work is something called convolutional neural networks. So these neural networks are generally developed for image recognition. It’s a part of this umbrella of machine learning and what that means is that this algorithm learns from data. We don’t specifically tell it what to look at in the data. So we show the algorithm all the correct examples. We show it a distorted image and then we tell it what the true answer is. And so by showing it a lot of examples, the networks can figure out what are the important key features in these things that are distinguishing the answers that we are looking for?
Interviewer: Anand Jagatia
And how well did the networks do? Did they do just as well as existing methods?
Interviewee: Yashar Hezaveh
Yeah they were doing absolutely amazingly. We were truly shocked how amazingly they were doing. They’re doing basically, as far as we know, very, very close to the position of traditional modelling methods.
Interviewer: Anand Jagatia
Is it much quicker than using the original methods that you were talking about?
Interviewee: Yashar Hezaveh
Yeah, so we were looking at something like more than ten million times improvement speed. And so here we can analyse a single gravitational lens in probably a hundredth of a second.
Interviewer: Anand Jagatia
Given that these networks are so much quicker, what could using them allow researchers to do?
Interviewee: Yashar Hezaveh
The first thing is that in the next decade we’re expecting to have thousands of new gravitational lenses that will be discovered. There are lots of interesting surveys that are coming up and so analysing these lenses with traditional methods would be absolutely difficult and so with neural networks we can just pass the data and within a few seconds have answers for all of them to teach us many things about the universe. The second thing has to do with complex models. So even for a single gravitational lens right now, sometimes these gravitational lenses are not as simple as what we said in this work. Sometimes clusters have hundreds of galaxies in them and all of them together are a distorting the image of the background galaxy. For these things, modelling a single lens gets immensely difficult, so people can spend years on the analysis of a single gravitational lens. So my hope is that with these neural networks we can actually model much more complex density structures.
Interviewer: Adam Levy
That was Yashar Hezaveh speaking to Anand Jagatia. Find the study in the usual place.
Interviewer: Shamini Bundell
Now it’s time for the News Chat and this week we’re talking again about a big story from a few weeks ago on gene-editing in human embryos. It was research that caused quite a big splash at the time but since then there’s been some debate about the claims made in the paper. Ewen Callaway has written a news piece about this and he joins me in the studio now. So Ewen, since it was published there have been some disagreements from some scientists about the results.
Interviewee: Ewen Callaway
Yes there have been some disagreements or some lack of clarity on some issues. Maybe I should refresh our listeners on what the paper was and what it showed. It was a landmark paper. That much is clear. It wasn’t the first report to use CRISPR Cas-9 genome editing in human embryos but it was by far the most successful and they were able to correct in human embryos this mutation linked to sudden cardiac death and what they reported was quite curious, I’d say, because normally the way you’d try and correct mutations is by – you introduce your own version of the gene you want to insert, right, so you put in a synthetic version. But normally you’d think it would be a healthy version of this gene. And the researchers found that that’s not actually how correction occurred. They found that correction occurred by copying the healthy version in the egg into the sperm which had the mutant version. So that was quite a surprising finding and quite an important one.
Interviewer: Shamini Bundell
When we were chatting about the paper we were talking about designer babies and the sort of controversial issues. This sort of surprising science finding that the maternal copy of the gene was being used as a kind of template to fix the broken version was just of minor interest but now that’s the bit that’s being called into question.
Interviewee: Ewen Callaway
Some scientists have concerns or have questions about whether Mitalipov’s team really did as they said. Did they prove that the maternal, healthy version of this gene was copied onto the sperm, onto the paternal version, to fix it? That’s what this is all about.
Interviewer: Shamini Bundell
So what’s their alternative theory for Mitalipov’s team’s results?
Interviewee: Ewen Callaway
They point out that okay, so when sperm fertilises an egg their genomes – so that you’ve got the sperm genome, the paternal genome, and the egg genome – they’re enshrouded in membranes on opposite sides of an egg cell. So they were saying when possibly can genome-editing occur such that maternal version is copied and replaced and it replaces the paternal version? So instead they’re suggesting that perhaps one of two things might have happened: sometimes eggs can self-fertilise. It’s called parthenogenesis; it happens in the wild, in some sharks, some other animals and you can induce it to occur in the lab in human embryos. They couldn’t develop into human beings but if you had a situation in which the egg self-fertilised and just kind of ignored the paternal genome it would look like you had two maternal copies of this gene. That’s one hypothesis. The other one is that maybe a huge chunk of DNA was removed from these embryos and they just couldn’t detect it because they weren’t using an appropriate genetics assay so they only detected the mother’s copy and they said, ‘aha!’ there’s no mutant copy here, we fixed it.
Interviewer: Shamini Bundell
Because they were trying to replace the mutant copy with their own version but when they looked their own version wasn’t there but the maternal copy was. So I guess the idea is maybe the maternal copy was the only thing they saw because they didn’t see the absence of the mutant maternal copy because it had been taken out.
Interviewee: Ewen Callaway
Exactly. That’s one possibility.
Interviewer: Shamini Bundell
And this isn’t just comments. They’ve written up a full paper. It’s on the BioRxiv preprint server but it hasn’t been peer reviewed or published yet?
Interviewee: Ewen Callaway
No, it was posted to the BioRxiv reprint server and then Mitalipov sent a statement saying basically he thought this was speculation but they say they’re preparing a response to explain why they believe these critics aren’t correct. I guess we’ll just have to stay tuned for that really.
Interviewer: Shamini Bundell
So we’ll have to have you back on to keep following that story. Thank you very much Ewen. And you can find Ewen’s News piece at nature.com/news. And for our original report you can listen to the Nature Podcastfor August 3rd.
Interviewer: Adam Levy
That’s all we’ve got time for this week but if you want to stay in touch, do make sure to follow us on Twitter. I’m @ClimateAdam.
Interviewer: Shamini Bundell
And I’m @SBundell.
Interviewer: Adam Levy
That’s imaginative.
Interviewer: Shamini Bundell
Thank you. Similarly as imaginative is Nature News. Find that @NatureNews. I’m Shamini Bundell.
Interviewer: Adam Levy
And I’m Adam Levy
[Jingle]