![]() |
| Click here and see if you can find this guy. Hint: he's smaller than the average rooster. |
Friday, May 31, 2013
Golden eggs
I just got back from visiting my lovely lady counterpart in Germany. In lieu of describing the whole trip, here is a photo of a gold-plated silver chicken.
Many, many more photos, presented with no context at all, are present here. Nope, that link is no longer operational. Photos available upon request.
Thursday, May 16, 2013
Do you have a second?
Every 15 minutes, someone averages a sum of events over an arbitrary period of time. No, I don't have a citation for that. It's standard practice to assume that frequencies are infrequently constant yet are much easier to understand if we can apply some kind of linear assumptions to them. That's perfectly fine, especially when we don't really know when the event in question is more or less likely to happen. This approach is used in science and medical journalism with the intent of breaking down impossibly huge numbers (i.e., 50 thousand deaths due to Ocelot Fever) into something more human-readable.
The problem arises when the level of granularity imposed by averaging over time obfuscates the reasons why the events happen at all. If MADD tells us that "In 2011, 9,878 people died in drunk driving crashes - one every 53 minutes", what do we really learn? That single number, (9,878 deaths / 53 minutes) is so specific that it disregards critical time-related factors like the weekend.*
I was planning to do an XKCD-inspired bit of Googling** and search for results of a few arbitrary frequencies to find out what happens at those times, every time. The results are too ghastly to share as most of the assumed events involve death, dismemberment, or assault. So, for the sake of diversity, I'll pull a few numbers from everysecond.info instead:
All I'm saying is that averaging massive numbers over time is misleading at best and dangerously myopic at worst. These are situations best handled by probabilities, not solid quantities and linear relationships. Averaging such an immense quantity over such a small period of time distorts our understanding of both quantities.
*This paper's thesis in brief: young people like to drink for fun on the weekends. To be fair, the study subjects were Swiss rather than the usual Americans. I'll refrain from griping about social science research for now.
**I have this nagging suspicion that this actually was an XKCD piece at some point. If so, go read that again too. It was probably pretty entertaining.
***250 kg of methane is 375 thousand liters, if that helps.
The problem arises when the level of granularity imposed by averaging over time obfuscates the reasons why the events happen at all. If MADD tells us that "In 2011, 9,878 people died in drunk driving crashes - one every 53 minutes", what do we really learn? That single number, (9,878 deaths / 53 minutes) is so specific that it disregards critical time-related factors like the weekend.*
I was planning to do an XKCD-inspired bit of Googling** and search for results of a few arbitrary frequencies to find out what happens at those times, every time. The results are too ghastly to share as most of the assumed events involve death, dismemberment, or assault. So, for the sake of diversity, I'll pull a few numbers from everysecond.info instead:
- Every second, Johnny Depp makes $2.92. The guy makes $92 million a year. Economists do use hourly pay or yearly salary to estimate how much an individual's time is worth, such that any period of time is "worth" however much they would have earned had they spent it working for pay. This can lead to some rather ludicrous estimates when inappropriately applied. We know Johnny Depp isn't actually working continuously despite his numerous sources of income. Even so, a number like $2.92/sec doesn't really provide us with any context to Mr. Depp's economic situation.
- Every second, 194 videos are watched on Myspace. What a perfect example! The statistic is from 2009. Some big changes have happened since then. Even so, it's fun to imagine Myspace users draining what's left of their attenuated attention spans on single-second videos. Hundreds of them every second!
- Every second, cows emit 250 kgs of methane in the United States. It's already difficult to imagine what a single kilogram of methane looks like,*** much less how what happens to it when the next second's round of methane arrives. Even so, cows don't continuously emit methane and they may emit more or less of it depending on what they're eating. Those factors can't even be considered when we break things down into seconds or even hours. If I observe a cow swish its tail three hundred times in an hour, perhaps I can safely claim 5 tail swishes per minute. Expanding the observation to a thousand cows over the course of months or years would render per-minute or per-second estimates useless without a greater knowledge of the relationship between cow tails and time.
All I'm saying is that averaging massive numbers over time is misleading at best and dangerously myopic at worst. These are situations best handled by probabilities, not solid quantities and linear relationships. Averaging such an immense quantity over such a small period of time distorts our understanding of both quantities.
*This paper's thesis in brief: young people like to drink for fun on the weekends. To be fair, the study subjects were Swiss rather than the usual Americans. I'll refrain from griping about social science research for now.
**I have this nagging suspicion that this actually was an XKCD piece at some point. If so, go read that again too. It was probably pretty entertaining.
***250 kg of methane is 375 thousand liters, if that helps.
Wednesday, May 08, 2013
Get well soon
Here is a use for one of the strange but inevitable results of modern society: singing greeting cards. The materials in one of those cards, plus a cheap resistor and capacitor, are enough to assemble a perfectly usable pulse sensor. The original paper suggests that this off-the-shelf solution could be used in an educational setting.
A little Googling shows that a number of pulse sensors are available or can be made for all kinds of platforms.
- This one is intended for use with Arduino. Total cost is ~$25, not counting whatever it's attached to.
- For comparison, this one is a more clinical model. I can only imagine how much it costs.
- This setup uses an optical approach instead of a piezoelectric one. It seems like overkill, possibly because the signal requires a lot of amplification.
I'm imagining an art project in which a viewer's heart rate changes the intensity of lights in a room or turns certain appliances (i.e., a fan or a radio) on or off at certain rate thresholds. Even something like a cheap knock-off of this bike helmet seems entirely feasible.
Thursday, May 02, 2013
More like "Nope! Share, Fool!"
It's worrisome to see politics bleed into science. It's even more alarming to see sources of scientific funding dry up because they don't fit a specific agenda. The National Science Foundation, under directive from Congress, recently cut off funding for political science research except for that "promoting national security or the economic interests of the United States." It's a rather general guideline for a specific field. It also reveals how legislators feel about the role of any science and when it deserves to be funded: we want results, we want them now, and we want we certainly don't need any context.
The NSF spent more than $7 billion last year. It's where 20 percent of the money for federally-funded research in the US comes from. That total goes a long way and contributes to numerous fields, from engineering to education. It's only a matter of time before some Congressman decides an entire field of research doesn't need federal funding at all.
The NSF spent more than $7 billion last year. It's where 20 percent of the money for federally-funded research in the US comes from. That total goes a long way and contributes to numerous fields, from engineering to education. It's only a matter of time before some Congressman decides an entire field of research doesn't need federal funding at all.
Tuesday, April 30, 2013
There aren't as many videos about yawning out there as you might think.
This Slate article about a supposed phenomenon known as ASMR is a few months old but it's still an interesting subject*. ASMR, or the Autonomous Sensory Meridian Response** is a rather general term for a collection of subtle but pleasurable psychological and physical reactions to what appears to be gentle or indirect stimuli. It can reportedly be triggered by gentle whispering, light touching, or even just the implication of receiving care from another person. Crinkling food wrappers, sorting silverware, or examining groceries also seem to produce ASMR for some individuals.
I don't experience this phenomenon. I generally remain skeptical about unexplained phenomena, too, especially when they're endemic to the Internet. Despite my skepticism I can't help but wonder if ASMR is really just a way to collectively codify a set of real psychological phenomena previously considered too minor to observe or too difficult to quantify. Everyday life is rich in dull but enjoyable moments. Perhaps mass communication just offers them in a more concentrated format, much like it does with news. Information overload is a real issue, too, if not a psychological one. It's the result of an inability to distinguish signal from noise; perhaps ASMR is pleasurable because it requires observation of a small signal (like Bob Ross' whispered instructions, for instance) in a low-noise background. It's relaxing because there isn't a lot of stimulus competing for the recipient's attention but it's stimulating because it's direct.
What's considered ASMR may truly be a collection of phenomena all related to the same set of primarily audiovisual stimuli. I think it's safe to suggest that the limbic system is involved and that much of the reaction is subconscious. It's reported to be not quite but almost orgasmic, much like yawning or sneezing are sometimes described. The Slate article above also suggests some potentially Freudian parent-child reaction but I think that may just be confirmation bias; Youtube participation may skew toward female users.
It would be interesting to see what range of stimuli repeatedly produce ASMRs in those who feel them. Do gently-worded threats do it? What about calming medical interviews in foreign languages?
*This sentence originally ended "...but it's been making the rounds again." That's how the last entry begins and it's kind of an embarrassing phrase to use anyway. Calling attention to it kinda defeats the purpose of editing but now it's not redundant, at least.
**It's really kind of a curious name for the phenomena. I'd generally expect unexplained psychological responses to acquire more pseudo-scholastic and less clinical names, i.e. Morgellons. Perhaps that's one distinction between a pleasurable unexplained phenomenon and a collective delusion.
I don't experience this phenomenon. I generally remain skeptical about unexplained phenomena, too, especially when they're endemic to the Internet. Despite my skepticism I can't help but wonder if ASMR is really just a way to collectively codify a set of real psychological phenomena previously considered too minor to observe or too difficult to quantify. Everyday life is rich in dull but enjoyable moments. Perhaps mass communication just offers them in a more concentrated format, much like it does with news. Information overload is a real issue, too, if not a psychological one. It's the result of an inability to distinguish signal from noise; perhaps ASMR is pleasurable because it requires observation of a small signal (like Bob Ross' whispered instructions, for instance) in a low-noise background. It's relaxing because there isn't a lot of stimulus competing for the recipient's attention but it's stimulating because it's direct.
What's considered ASMR may truly be a collection of phenomena all related to the same set of primarily audiovisual stimuli. I think it's safe to suggest that the limbic system is involved and that much of the reaction is subconscious. It's reported to be not quite but almost orgasmic, much like yawning or sneezing are sometimes described. The Slate article above also suggests some potentially Freudian parent-child reaction but I think that may just be confirmation bias; Youtube participation may skew toward female users.
It would be interesting to see what range of stimuli repeatedly produce ASMRs in those who feel them. Do gently-worded threats do it? What about calming medical interviews in foreign languages?
*This sentence originally ended "...but it's been making the rounds again." That's how the last entry begins and it's kind of an embarrassing phrase to use anyway. Calling attention to it kinda defeats the purpose of editing but now it's not redundant, at least.
**It's really kind of a curious name for the phenomena. I'd generally expect unexplained psychological responses to acquire more pseudo-scholastic and less clinical names, i.e. Morgellons. Perhaps that's one distinction between a pleasurable unexplained phenomenon and a collective delusion.
Sunday, April 21, 2013
Hunka burnin' love
There's a web game called No One Has To Die that's been making the gaming-blog rounds. I don't intend this to be a gaming blog per se but this particular game is really quite good. Without giving too much away, it an excellent example of how even the smallest dose of interactivity can turn a fairly average story into a multifaceted one by virtue of player choice. The venerable Choose Your Own Adventure books did that too, of course, but aren't those really games, too? Either way, No One Has To Die isn't even that much more fun than flipping around a paperback to avoid the "bad" endings, but in this case it's clearly intentional. Perhaps most games can be reduced to repeating different permutations of the same actions until the desired result is reached.
Friday, April 19, 2013
As cheap as free?
It now costs about $3000 to $5000 to sequence a human genome. That cost is 10,000x less than the same effort had required just 10 years ago. Obviously we can't get the technology to be much cheaper, right? The sequencing machinery itself is growing cheaper: as of last year, a decent next-generation sequencer could be purchased for $80,000 to $120,000 or so, down from half-million dollar models from just a few years ago. The limiting factor here may be how the machines are intended for research purposes. A well-funded lab can certainly afford to spend hundreds of thousands of dollars on a single machine, to say nothing of the reagents required to actually get data out of it.
Ubiquitous DNA sequencing has the potential to upend nearly everything we know about personalized medicine, but only if it's not cost-prohibitive. Basic molecular biological techniques certainly have applications in clinical or diagnostic environments. Sequencing could really go beyond that and become relevant to everyday folks. It could just be a matter of having a streamlined, automated system, with samples sent elsewhere. A recent start-up called uBiome offers to characterize the microbiome of just about anything you'd like for less than $100. They aren't specific about how they got their costs so low, but I'm assuming they're just doing the usual 16S rRNA amplification and some quick next-gen sequencing. That's one way to create demand for incredibly cheap sequencing technologies.
Let's wax futurist about the potential of super-cheap, publicly-available nucleotide sequencing:
Ubiquitous DNA sequencing has the potential to upend nearly everything we know about personalized medicine, but only if it's not cost-prohibitive. Basic molecular biological techniques certainly have applications in clinical or diagnostic environments. Sequencing could really go beyond that and become relevant to everyday folks. It could just be a matter of having a streamlined, automated system, with samples sent elsewhere. A recent start-up called uBiome offers to characterize the microbiome of just about anything you'd like for less than $100. They aren't specific about how they got their costs so low, but I'm assuming they're just doing the usual 16S rRNA amplification and some quick next-gen sequencing. That's one way to create demand for incredibly cheap sequencing technologies.
Let's wax futurist about the potential of super-cheap, publicly-available nucleotide sequencing:
- Breakfast cereal manufacturers can offer free microbiome sequencing in every box of Cap'n Bran Flakes to show how their product might contribute to a healthy gut.
- Vending machines can offer genome screening on demand to screen for potential genetic maladies (but do people ever use those drugstore blood pressure screening machines? Are they secretly some kind of Scientologist apparatus?)
- Speaking of drugstores, pharmacies could do quick screens for genetic predisposition to adverse drug reactions before they actually dispense said drugs.
- Labs could spend money on more practical things, like this electronic ice bucket. Oh yeah.
Monday, April 15, 2013
Virulence Factors is the punk band that never was*
It seems like every time I hit up Google for a quick answer to a science question I find a new database. In this case, the question was "does the Vibrio pathogenicity island produce an infective phage particle on its own?" It's really just a question best answered by a review article rather than a database, but I found a nice map of the pathogenicity island in question. Very helpful, plus this VFDB site may be helpful in the future despite, sadly, appearing inactive since 2008.**
*Virulence Factor is apparently a band from Chicago. They're kind of Evanescence-metal.Unlike the above database, Virulence Factor remains active.
**Oh, wait, no! There's a new version of VFDB.
*Virulence Factor is apparently a band from Chicago. They're kind of Evanescence-metal.
**Oh, wait, no! There's a new version of VFDB.
Thursday, April 11, 2013
A membrane protein specific to Staph?
So I was poking around in some gene orthology databases and found a family of uncharacterized proteins conserved in Staphylococcus. It's at least in S. aureus COL. Though the protein family is largely undescribed, some annotations seem to suggest it's a M50B family metalloendopeptidase.
Wednesday, April 10, 2013
Pretty pictures
| No radiation, but it is diatomic |
The image is from Anne Weston of the London Research Institute of Cancer Research UK and is a 2012 Wellcome Image Award winner.
Sunday, April 07, 2013
Babies on strike
Found while looking for tuberculosis-related public health posters:
This is a poster from the Library of Congress archives. Perhaps babies just aren't well-organized.
| Infants Local 23 |
Saturday, April 06, 2013
Here's a pretty neat example of technological time travel: media preservation folks at Indiana University found gramophone records printed into books from the late 19th century. The records aren't solid media. They're just ink-based reproductions of the grooves found on records. From these grooves we can produce waveform images and from the images we can produce audio. The original authors of these reproduced records explicitly intended their work to be accessible in the future, though they clearly didn't foresee technology like scanners or audio editing software.
There are other, earlier examples of hand-engraved waveforms. Here's one from 1806, mentioned in the link above but looped by me just for fun:
There are other, earlier examples of hand-engraved waveforms. Here's one from 1806, mentioned in the link above but looped by me just for fun:
Thursday, April 04, 2013
I listened to this TED talk by Allan Savory today. TED talks are usually rich in big ideas and poor in specifics. This one is certainly no exception. Even so, it illustrates one potential example of an established phenomenon proving to be somewhat less than the absolute truth.
Savory explains how global climate change isn't just due to an overabundance of burning carbon sources. It's also due to rampant desertification. Many former grasslands have, over the course of the last century or so, lost most of their natural plant life and groundcover, exposing the soil to the elements and increasing water loss by evaporation. The area no longer serves as carbon storage. Savory proposes that the only solution (and here's where I start to get skeptical, at least due to his claims of the "only solution") is to mimic the effects of natural animal movement patterns with livestock. Large herds of cattle usually don't stay in one place for long: they graze for a while and move on to the proverbial greener pastures. Some success has been had in planning the grazing patterns of domestic livestock such that they keep moving through arid areas and stimulate plant growth rather than just consuming it all. The improved plant growth should prevent or even reverse large swaths of desert and allow for enough carbon storage to reduce atmospheric carbon to pre-industrial levels.
It would be nice to think that reversing carbon-based climate change is as easy as shuffling around livestock. I worry about how well the theory scales up; what assumptions are being made about how much desert land can really be reverted to grassland? Do these plans require more livestock than we currently have? What's the potential impact on methane levels? Allan Savory has been talking about holistic resource management for decades, but some actual models of its potential impact on climate change would be nice to see. I presume they exist somewhere. (There's plenty of climate-bloggery about Savory's hypotheses out there but much of it looks less than open-minded.)
Savory explains how global climate change isn't just due to an overabundance of burning carbon sources. It's also due to rampant desertification. Many former grasslands have, over the course of the last century or so, lost most of their natural plant life and groundcover, exposing the soil to the elements and increasing water loss by evaporation. The area no longer serves as carbon storage. Savory proposes that the only solution (and here's where I start to get skeptical, at least due to his claims of the "only solution") is to mimic the effects of natural animal movement patterns with livestock. Large herds of cattle usually don't stay in one place for long: they graze for a while and move on to the proverbial greener pastures. Some success has been had in planning the grazing patterns of domestic livestock such that they keep moving through arid areas and stimulate plant growth rather than just consuming it all. The improved plant growth should prevent or even reverse large swaths of desert and allow for enough carbon storage to reduce atmospheric carbon to pre-industrial levels.
It would be nice to think that reversing carbon-based climate change is as easy as shuffling around livestock. I worry about how well the theory scales up; what assumptions are being made about how much desert land can really be reverted to grassland? Do these plans require more livestock than we currently have? What's the potential impact on methane levels? Allan Savory has been talking about holistic resource management for decades, but some actual models of its potential impact on climate change would be nice to see. I presume they exist somewhere. (There's plenty of climate-bloggery about Savory's hypotheses out there but much of it looks less than open-minded.)
Wednesday, April 03, 2013
Friday, March 29, 2013
![]() |
| Guess who's BCKDK (this is BCKDK, or at least its structure). Image by Wikipedia user Emw. |
Speaking of awkward acronyms: BCKDK.
It's branched-chain alpha-ketoacid dehydrogenase kinase, an enzyme critical to amino acid metabolism in humans. Defects in BCKDK are linked with some forms of autism.
Creating new methods is great for three reasons:
Citation:
Moreno, D., Neller, J., Kestler, H., Kraus, J., Dünkler, A., & Johnsson, N. (2013). A fluorescent reporter for mapping cellular protein-protein interactions in time and space Molecular Systems Biology, 9 DOI: 10.1038/msb.2013.3
- It allows old phenomena to be explored in new ways.
- It provides other researchers with more options.
- It permits creation of ludicrous acronyms.
A recent paper by Moreno et al. in Molecular Systems Biology is an example of cases 1 and 3. Perhaps it will someday be useful to researchers studying protein-protein interactions, but for now, SPLIFF looks like a good way to visualize interactions found by other assays (i.e., yeast two-hybrid).
But I'm getting ahead of myself. This is a method called SPLIFF. It's about lighting up cells. Somebody got a little carried away with that acronym.
SPLIFF is a Split-Ubiquitin method and indicates protein-protein interactions within cells though the ratio fluorescence produced by two auto-fluorescent reporter proteins. Assays for protein-protein interactions usually use a growth phenotype or a colorimetric reporter to indicate interactions, so SPLIFF may offer more clearly quantifiable, temporally-sensitive results. The authors also show how SPLIFF, like other fluorescence-based methods, can localize interactions within cells. The protein interactions are reportedly detected instantaneously. Hopefully this method can weed out some of the false-positives in two-hybrid interaction screens.
Citation:
Moreno, D., Neller, J., Kestler, H., Kraus, J., Dünkler, A., & Johnsson, N. (2013). A fluorescent reporter for mapping cellular protein-protein interactions in time and space Molecular Systems Biology, 9 DOI: 10.1038/msb.2013.3
Tuesday, March 26, 2013
Here's an interesting example of emergent properties.
In short, it's essentially a topology-based approach to solving a classical optimization problem. Quite novel, really, though it's not clear if the solution offers a whole lot of benefit over brute force calculation in terms of saved computation time. The paper is here.
What struck me as particularly interesting was the authors' proposal to physically implement their "blob" model:
In short, it's essentially a topology-based approach to solving a classical optimization problem. Quite novel, really, though it's not clear if the solution offers a whole lot of benefit over brute force calculation in terms of saved computation time. The paper is here.
What struck me as particularly interesting was the authors' proposal to physically implement their "blob" model:
...it would be satisfying if this virtual material could be implemented and embodied in a real physical substrate with the desired physical (for example visco-elastic, free energy minimisation) properties.
What would this accomplish? I'm not saying that it wouldn't be a neat, cheap engineering project to build an elastic blob capable of solving traveling salesman problems, it just doesn't seem like any kind of a job for real-world models.
Jeff Jones, & Andrew Adamatzky (2013). Computation of the Travelling Salesman Problem by a Shrinking Blob arXiv arXiv: 1303.4969v2
Monday, March 25, 2013
Here's a paper dealing with a frequently worrisome issue:
Shrestha, P. M., Nevin, K. P., Shrestha, M., & Lovley, D. R. (2013). When Is a Microbial Culture “Pure”? Persistent Cryptic Contaminant Escapes Detection Even with Deep Genome Sequencing. mBio, 4(2). doi:10.1128/mBio.00591-12
Microbiologists, like most biologists, like to have neatly organized categories and taxonomies available at all times. We work with individual species, specific strains of those species, and often mutants of those strains. If we grow our chosen organisms and they don't behave as expected, we generally assume that contamination occurred somewhere along the line or that a new mutation transpired. The usual preventative measure is simply to ensure that cultures start from single cells of the desired strain.
Shrestha et al. seem to have had some issues with culture contamination. This group works with Geobacter sulfurreducens strain DL1 in studies of microbial fuel cells. As they were attempting to isolate G. sulfurreducens strains with mutations enabling them to be more electrically conductive, they managed to isolate a strain later named KN400. This strain was great at conducting current but clearly wasn't just a mutant of DL1: sequencing showed that the new strain likely differed by more than 27,000 SNP's, or far more mutations than any lab-based methods of selection were likely to create.
Microbiologists, like most biologists, like to have neatly organized categories and taxonomies available at all times. We work with individual species, specific strains of those species, and often mutants of those strains. If we grow our chosen organisms and they don't behave as expected, we generally assume that contamination occurred somewhere along the line or that a new mutation transpired. The usual preventative measure is simply to ensure that cultures start from single cells of the desired strain.
Shrestha et al. seem to have had some issues with culture contamination. This group works with Geobacter sulfurreducens strain DL1 in studies of microbial fuel cells. As they were attempting to isolate G. sulfurreducens strains with mutations enabling them to be more electrically conductive, they managed to isolate a strain later named KN400. This strain was great at conducting current but clearly wasn't just a mutant of DL1: sequencing showed that the new strain likely differed by more than 27,000 SNP's, or far more mutations than any lab-based methods of selection were likely to create.
The authors sought to clarify the origins of G. sulfurreducens strain KN400. Next-gen sequencing showed that a sequence specific to KN400 could be found in the original DL1 cultures, though out of more than 107 sequence copies, fewer than 300 were the KN400 sequence. The new strain wasn't a new mutant at all. Subsequent growth studies found that even repeated rounds of restreaking cells on solid medium still yielded colonies with minute but detectable levels of the cryptic KN400 strain. Shrestha et al. suggest that their intense electrical conductivity-based selective pressures may have been the only reason they were able to isolate KN400 at all.
It's still not clear how G. sulfurreducens KN400 lives alongside the DL1 strain so well. My guess is that KN400 is actually an instance of a separate but phylogenetically related chromosome copy acting as a mobile genetic element. Cultures of the DL1 strain consistently maintain KN400 cells at low levels, so some variety of fitness benefit must be conferred to the entire culture or we'd expect the minority cells to just get selected against. There's clearly something strange going on with this bacterial species but it may be an ongoing process.
Friday, March 22, 2013
Gatsogiannis, C., Lang, A. E., Meusch, D., Pfaumann, V., Hofnagel, O., Benz, R., Aktories, K., et al. (2013). A syringe-like injection mechanism in Photorhabdus luminescens toxins. Nature, advance on. Nature Publishing Group, a division of Macmillan Publishers Limited. All Rights Reserved. doi:10.1038/nature11987
Didn't read this one yet, but the abstract contains the phrase "vuvuzela-shaped channel" so it should be interesting.
Didn't read this one yet, but the abstract contains the phrase "vuvuzela-shaped channel" so it should be interesting.
Subscribe to:
Posts
(
Atom
)





