May 31, 2016 at 12:29AM

Today I Learned: 1) Fruit flies can jump! It's hard to tell if they jump in general, because they move so darned fast, but I watched a bunch of wingless fruit flies today, and they were definitely jumping around between things and off of things. 2) Characters like "µ" can really get you in trouble in Python (by which I mean they'll cause you to throw an error, which I guess isn't *that* bad in the grand scheme of things). Today I learned how to put non-ASCII characters in a Python string -- just put a "u" before the string, like this: u'GFP (µM)'. This causes Python to interpret the string as UTF-8 instead of ASCII, which lets it recognize pretty much any character. ...at least, that's what you do in Python 2. Not sure about 3. 3) Want a convincing starfield for a picture background? Try this -- leave some blank white paper on a table for a while, then scan it (preferably with an old, dusty scanner). You'll get a white field with a bunch of flecks of dust. Invert the image, giving yourself a black image with a bunch of white flecks. Add some blur to make the flecks look more like stars. Voila! Starfield! ImageJ is good for this kind of image manipulation. I hear paint.NET is also pretty good? Thanks to Sarah Seid for teaching me this method.

May 30, 2016 at 02:29AM

Today I Learned: 1) A biohacker group called Science for the Masses (they look pretty interesting -- might be worth checking them out for other purposes) has prototyped an eye drop that temporarily increases night vision capabilities. It's based on a chlorophyll-like molecule called chlorin e6, or Ce6, which absorbs red light. According to a 2007 paper, When mice were administered Ce6 (I'm not sure how), they gained the ability to see red light better under low-light conditions. The current hypothesis, from what I gather, is that Ce6 couples to rhodopsins, essentially giving them greater sensitivity to red light. The biohackers dissolved a bunch of Ce6 in salt water along with insulin and DMSO, and pipetted a few drops of the mix into one of their eyes. I'm not sure what the insulin is supposed to do, but DMSO is a chemical that's really good at moving other chemicals through barriers like clothes, gloves, and cell membranes, so it's probably there to help Ce6 diffuse into the retina. The results were promising but not conclusive -- the guy who tried the drops identified objcts at range in the dark much better than four control subjects who didn't receive the drops, but the sample size was small, the study wasn't blind, and there was no correction for baseline differences in identification. In other words, it was sloppily done, so the results aren't solidly convincing. Also, DON'T try this at home, because there's some reason to think administering Ce6 to your eyeball could be dangerous to your eyesight, especially when used repeatedly. Ce6 is currently used medically... to kill tumor cells. Specifically, when it absorbs red light, it dumps out the energy in the form of reactive oxygen species (ROS). ROSs are toxic to cells at high concentration. You can inject a bunch of Ce6 around a tumor and hit it with a red laser or two, producing a whole bunch of ROSs locally around the tumor, hopefully killing it. If you put it in your *eyeball*, then there's a decent chance that if you were exposed to bright light*, you'd have a similar effect, but in your retinas, which you generally don't want to kill. On the other hand, if Ce6 really does complex to rhodopsins, I suspect they're dumping their energy into the rhodopsins instead of producing ROSs... but I wouldn't want to bet my eyesight on that, or on *all* (or even most) of the Ce6 actually getting bound to rhodopsin. For a more thorough review of this particular experiment, check out http://ift.tt/25sFB7C. * The Science for the Masses hacker who tried this put on dark contacts during the test and protected his eyes from light after receiving the drops, which is both a sensible safety precaution and another probable confounder of the experiment if the other subjects didn't. 2) There are apparently decent, controlled studies on the effectiveness of echinacea herb, garlic, vitamin C, and zinc. The results are, respectively: doesn't help colds; probably helps prevent colds and reduce severity, if taken daily, but has side effects; definitely, *definitely* doesn't help colds* unless, maybe, *possibly*, you are undergoing strenuous physical training at high altitudes, and ; might improve outcomes, not enough data to say if it's preventative, and probalby isn't worth the side effects. There have *not* been controlled studies on the effectiveness of either chicken soup or Airborne™ on cold recovery. The company that makes Airborne™ used to claim that it was backed up by a study, but the "study" in question was, in fact, run by two non-doctor, non-scientist fellows, who put together their two-man company specifically to run that one trial. That particular claim has been removed following a class-action lawsuit. Oh, and Airborne™ is *not* approved by the FDA (as a supplement, it doesn't need to be -- this is an important fact, but the way: there are essentially no safety or efficacy requirements other than "it isn't acutely toxic" and "it doesn't claim to do anything medically" for dietary supplements. A company could sell flavored sugar pills, and as long as they labeled it accurately and tell you that it doesn't cure anything, they can claim whatever they want about its effects and sell it to you.) * or, as far as I know, anything else unless you're about to get scurvey -- seriously, vitamin C as a medicine was a crazy idea popularized by Linus Pauling that has *never* panned out clinically, but that loads of companies have latched onto to sell you supplements. Don't bother unless you can't find a cheaper sugar pill. 3) I ran across this article (http://ift.tt/1slQk5m), a design article on Shovel Knight, a modern video game designed to essentially be what would have been developed today if development for the NES had never stopped. It's a cool, quick look at some of the technical limitations of the NES, how programmers of the NES era dealt with those limitations, and which ones the developers decided to abide by which ones they decided to break, and why. Example: large sprites were very difficult to render on the NES, so games like Megaman drew bosses and other large enemies in the background layer rather than as a sprite (or multiple sprites). The only problem is that bosses need a lot of animations, and it's a pain to make a different background for every single animation frame. That's why NES game bosses were usually drawn on black or otherwise dark, uniform backgrounds. This was one place where Shovel Knight remains pretty faithful to the original. They do use sprites for bosses, because that's a heck of a lot easier to do as long as your system can support it, but they kept the dark backgrounds because "We thought that the black background with the huge boss always gave NES games a distinctive and epic feel, where the focus was just on you and your enemy".

May 30, 2016 at 02:29AM

Today I Learned: 1) A biohacker group called Science for the Masses (they look pretty interesting -- might be worth checking them out for other purposes) has prototyped an eye drop that temporarily increases night vision capabilities. It's based on a chlorophyll-like molecule called chlorin e6, or Ce6, which absorbs red light. According to a 2007 paper, When mice were administered Ce6 (I'm not sure how), they gained the ability to see red light better under low-light conditions. The current hypothesis, from what I gather, is that Ce6 couples to rhodopsins, essentially giving them greater sensitivity to red light. The biohackers dissolved a bunch of Ce6 in salt water along with insulin and DMSO, and pipetted a few drops of the mix into one of their eyes. I'm not sure what the insulin is supposed to do, but DMSO is a chemical that's really good at moving other chemicals through barriers like clothes, gloves, and cell membranes, so it's probably there to help Ce6 diffuse into the retina. The results were promising but not conclusive -- the guy who tried the drops identified objcts at range in the dark much better than four control subjects who didn't receive the drops, but the sample size was small, the study wasn't blind, and there was no correction for baseline differences in identification. In other words, it was sloppily done, so the results aren't solidly convincing. Also, DON'T try this at home, because there's some reason to think administering Ce6 to your eyeball could be dangerous to your eyesight, especially when used repeatedly. Ce6 is currently used medically... to kill tumor cells. Specifically, when it absorbs red light, it dumps out the energy in the form of reactive oxygen species (ROS). ROSs are toxic to cells at high concentration. You can inject a bunch of Ce6 around a tumor and hit it with a red laser or two, producing a whole bunch of ROSs locally around the tumor, hopefully killing it. If you put it in your *eyeball*, then there's a decent chance that if you were exposed to bright light*, you'd have a similar effect, but in your retinas, which you generally don't want to kill. On the other hand, if Ce6 really does complex to rhodopsins, I suspect they're dumping their energy into the rhodopsins instead of producing ROSs... but I wouldn't want to bet my eyesight on that, or on *all* (or even most) of the Ce6 actually getting bound to rhodopsin. For a more thorough review of this particular experiment, check out http://ift.tt/25sFB7C. * The Science for the Masses hacker who tried this put on dark contacts during the test and protected his eyes from light after receiving the drops, which is both a sensible safety precaution and another probable confounder of the experiment if the other subjects didn't. 2) There are apparently decent, controlled studies on the effectiveness of echinacea herb, garlic, vitamin C, and zinc. The results are, respectively: doesn't help colds; probably helps prevent colds and reduce severity, if taken daily, but has side effects; definitely, *definitely* doesn't help colds* unless, maybe, *possibly*, you are undergoing strenuous physical training at high altitudes, and ; might improve outcomes, not enough data to say if it's preventative, and probalby isn't worth the side effects. There have *not* been controlled studies on the effectiveness of either chicken soup or Airborne™ on cold recovery. The company that makes Airborne™ used to claim that it was backed up by a study, but the "study" in question was, in fact, run by two non-doctor, non-scientist fellows, who put together their two-man company specifically to run that one trial. That particular claim has been removed following a class-action lawsuit. Oh, and Airborne™ is *not* approved by the FDA (as a supplement, it doesn't need to be -- this is an important fact, but the way: there are essentially no safety or efficacy requirements other than "it isn't acutely toxic" and "it doesn't claim to do anything medically" for dietary supplements. A company could sell flavored sugar pills, and as long as they labeled it accurately and tell you that it doesn't cure anything, they can claim whatever they want about its effects and sell it to you.) * or, as far as I know, anything else unless you're about to get scurvey -- seriously, vitamin C as a medicine was a crazy idea popularized by Linus Pauling that has *never* panned out clinically, but that loads of companies have latched onto to sell you supplements. Don't bother unless you can't find a cheaper sugar pill. 3) I ran across this article (http://ift.tt/1slQk5m), a design article on Shovel Knight, a modern video game designed to essentially be what would have been developed today if development for the NES had never stopped. It's a cool, quick look at some of the technical limitations of the NES, how programmers of the NES era dealt with those limitations, and which ones the developers decided to abide by which ones they decided to break, and why. Example: large sprites were very difficult to render on the NES, so games like Megaman drew bosses and other large enemies in the background layer rather than as a sprite (or multiple sprites). The only problem is that bosses need a lot of animations, and it's a pain to make a different background for every single animation frame. That's why NES game bosses were usually drawn on black or otherwise dark, uniform backgrounds. This was one place where Shovel Knight remains pretty faithful to the original. They do use sprites for bosses, because that's a heck of a lot easier to do as long as your system can support it, but they kept the dark backgrounds because "We thought that the black background with the huge boss always gave NES games a distinctive and epic feel, where the focus was just on you and your enemy".

May 28, 2016 at 12:27AM

Today I Learned: 1) The most poisonous tree (not the most poisonous plant) is the Manchineel tree, native to Florida. Its fruit is toxic, and can kill (though usually it just gives a super nasty rash down its eater's throat -- it is, however, apparently sweet and quite tasty). Its sap is caustic, and will produce painful blisters on contact. If you get it in your eye, it will blind you (temporarily). Pretty much every single part of the tree will hurt you if you touch it. One might wonder why a tree would bear a fruit that's toxic. Many plants have fruit that's toxic (holly) or un-tasty (jalapeños) to mammals to keep them from eating it while still being edible to their main seed-spreading vector, birds. Manchineel fruit looks a bit big to be eaten by birds to me, but it might have its seeds spread by something else. Another proposal is that the Manchineel tree's fruits spread by water, but then one would wonder why they would bother making sugar-laden, pulpy fruit. On a side note, Manchineels are endangered. 2) Today I read about a rather remarkable system for tracing the lineages of cells in an organism. First, a word on what that means. Imagine a fertilized fish egg. It's one cell. It's going to eventually grow into a fish with somewhere between tens of millions and trillions of cells. Developmental biologists would really like to know how that cell and its descendents split to to eventually make up the fish. The combined history of all of the cells in the organism is called a lineage tree (or fate map), and the process of figuring out the lineage tree is generally called fate mapping. To fate-map an entire organism is... difficult, to say the least. A lab at Harvard has developed a system called GESTALT (Genome Editing of Synthetic Targetted Arrays for Linage Tracing) for doing just that, at least in a coarse-grained way. It uses... Cas9! Basically, the system uses an array of nine or ten repeated sequences, weakly targetable by a genome-editing version of Cas9 with a guide expressed as part of the construct, plus Cas9 itself. The whole thing is added to the organism's genome at a single-celled stage. As the ogranism grows and divides, Cas9 will occasionally "mark" one of the editing sites with an insertion or deletion. Once it's made an edit in a cell, that cell and *all of its descendents* will have the mutation. Once the organism is fully grown, you can take out single cells, sequence them at the array site, and figure out which ones are related by which mutations they share. Credit on finding out about this goes to Andy Halleran. Apparently he heard about this work while he was interviewing. This is why you should apply to grad schools, kids -- interviews are a fantastic way to travel the country and hear about all the best science before it gets published. * We actually have this for the nematode C. elegans, which has an extremely rigid lineage tree and a fixed number of cells (1031, in the case of the male). It was determined in the early 80s by watching a bunch of developing C. elegans very closely. 3) There are viscoelastic materials that, when compressed, spring back to *more* than their original size. Does bread dough do this? Thanks for teaching me this, Mengsha Gong!

May 27, 2016 at 04:02AM

Today I Learned: 1) I found myself, earlier today, on Monsanto's "Myths about Monsanto" webpage. This led me to look up some information about a rather famous case of Monsanto's, which I'd bet you've heard of but I'd also bet you don't know the full story about -- Monsanto vs. Percy Schmeiser. This is the farmer that Monsanto sued when some of Monsanto's GM strain drifted over into the farmer's field and contaminated his crop. At least, that's what happened according to anti-GMO activists and Fellows Posting On The Internet. It seems that's not quite what happened. The real story is more complicated, and a lot more interesting. Percy Schmeiser had a canola farm nearby to some farmers who grew Monsanto crops with resistance to glyphosate, better known as Roundup. One day, he noticed that some crops he'd sprayed with roundup to clear them out hadn't entirely died. Suspicious, Schmeiser cordoned off one of his fields and sprayed the whole thing with roundup. About half of it died. The other half was roundup-proof. So the first crazy thing about this incident is that fifty percent number. Schmeiser didn't just have a little contamination -- HALF OF HIS CROP was Monsanto-derived, through no fault of his own. This is kind of crazy, but to be perfectly clear, Monsanto *did not* sue him for having GM crops on his field without permission. They sued him for what he did next. What he did next was to harvest the remaining canola plants off of the Roundup-ed plot, which he now knew were Monsanto-strain canola plants. He stored the seeds from those crops, then planted them on a bunch of fields the next planting season. When Monsanto found out, they brought a suit against Schmeiser for knowingly and intentionally planting their strain without permission or lease. Schmeiser, in his defense, argued that since the seeds were on his property, through no illegal action of his own, they were his to use as he saw fit. Monsanto argued that since the seeds were Monsanto intellectual property and Schmeiser had willfully used them without permission, Schmeiser had violated Monsanto's intellectual property. The court ruled 5-4 in favor of Monsanto. The case has since been appealed multiple times, and Monsanto has won every time. So Monsanto isn't quite the ridiculously evil caricature that the usual story makes them out to be. They kind of have a point -- Schmeiser went through a fair amount of effort to isolate the GM strain in his fields (which he hadn't wanted in the first place, or at least not wanted to pay for) and then intentionally used them in his fields. Then again, it's really weird to me that Monsanto can say who does and doesn't use their seeds, even multiple generations after they were purchased. What do you think of this story? 2) While I'm on topic of Monsanto, today I learned that Monsanto does not sell, and has never sold, plants with a "terminator" gene of any kind that would make them sterile. They researched such a technology, and decided not to pursue it due to public outcry, despite the fact that it would a) probably make them a ton of money and b) prevent un-wanted spread of Monsanto's GMO strains. 3) It's been a while since I've read up what's going on in EVE Online*, so today I did that. The newest news in EVE Online is that there's been a massive turnover of the biggest player. A year ago, one of the bigger (player-controlled) factions was The Imperium, which apparently gained dominance through a combination of ruthlessness, sheer wealth, and cunning diplomacy. Not, however, through the actual ability to hold territory -- the leader of the Imperium, named "The Mittani" in-game, has plainly said that the Imperium was tremendously over-stretched for about half a year, but nobody dared challenge them because, fraknly, they *looked* unassailable. That changed about two months ago when a group of bankers calling themselves Moneybadger Coalition sent a war fleet into Imperium space. For the most part, the Imperium liquidated their considerable assets, pulled out with tons and tons of money, and left. There's still some fighting going on, but the Imperium is *definitely* not the dominant power in the Eve Online universe anymore. Moneybadger Coalition isn't just bold. They have a *lot* of in-universe money. One of the heads of Moneybadger Coalition has something like 11 trillion ISK in his in-game account, which is worth about $300,000 USD real-world. That's enough money to personally bankroll the largest battle EVE online has ever seen. How did he get so rich? Gambling. Specifically, he (and some other big EVE bankers) own a third-party website called "I Want ISK", in which players can gamble their EVE Online ISKs. The website, of course, skims off a profit. It's been an enormously successful venture, and frankly nobody in the game really knows how to counter that kind of cash flow. This also raises some ethical and legal issues... there's no age control on EVE online, or on I Want ISK, so the whole system might be subsidizing, for example, child gambling, or gambling in states where it would not normally be legal (and in fact, might not be). EVE Online continues to surprise me. *For the uninitiated, EVE Online is an MMORPG set in a spacefaring empire of the far future. It is most noted for its almost-completely-player-driven economy, the massively complex real-world, player-run organizations that run things, and the immense amounts of time required to play. It was created by Iclandic company CCP Games, partly to be a game and partly to be a controlled experimental setting for economics geeks. So far, it has done pretty well at both.

May 26, 2016 at 03:40AM

Today I Learned: 1) Snail mouths are beautiful! Example: http://ift.tt/27TpMsI On a similar note, have you ever looked inside a sea turtle's mouth? It's bizarre. http://ift.tt/1OOFgT8. Those spiky bits go quite far down the turtle's esophagus. Sea turtles have super-long, coiled-up esophagi, presumably because the jellyfish they largely subsist on aren't very nutrient-rich and take a while to digest, so they need to store a lot of them. The spikes are to keep any jellyfish from escaping when the turtle expells seawater from its stomach. Thanks for sending me down this rabbit hole, Sarah Seid! 2) This is something I already knew about, but every time I run across this I'm amazed anew... in class today, our professor showed us an analysis of some data from some sort of pulsatile gene in one of the university's labs. The data were really simple -- just a list of times between pulses of the gene. If you plot a histogram of those times, it looks pretty much like an exponential distribution -- lots of very short inter-spike times, with a long tail of larger spike times. The question we addressed was this: is the distribution of arrival times actually exponentially-distributed, or is it something that looks superficially like an exponential distribution? Why that question? Well, if the gene's pulses were random, uncorrelated events with equal probability of happening at any particular time, then their inter-pulse distribution would be exponentially distributed. If they're *not* exponentially distributed, it means there's something else going on. In this case, we compared three hypotheses -- the exponential distribution, which, as I said, you get if you assume that every pulse is random and independent of the others; a Weibull distribution, which is what you would get if the underlying process was random but became either more or less likely over time (kind of like the likelihood of equipment failure -- it's more or less random, but becomes more likely as the equipment gets older); and a double-exponential distribution, which would occur if there were two distinct states with different pulse rates that the cell could go into after each pulse. We very quickly went through a Bayesian analysis that gave us, in the end, a ratio of likleihoods between each of the different hypotheses. We also got the best-fit curve for each hypothesis, and compared each to the data. The exponential distribution fit pretty darned well for moderate interpulse times, but deviated noticably at small and large interpulse times. The Weibull distribution fit slightly better, but not much. The double-exponential distribution fit the front half perfectly, and was about as accurate as the other two at high interpulse times. So, which hypothesis was most likely? The double-exponential. Sure, fine, whatever -- what startled me was *how much more likely* the double-exponential was. If you assume before the experiment that the double-exponential and exponential distributions were more or less equally likely, and after accounting for differences in model complexity, the double-exponential was something like TEN to the ONE HUNDRED AND FIFTY FOUR times more likely than the exponential. To put that number just a tiny little bit into perspective, let me write that out, with some benchmarks annotated: 10,000,000,000,000,000,000,000,000,000,000,000∞,000,000,000,000,000,000,000,000,000,000,000,000,00¥0,000◊,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,00†0,000,0°00,00§0,000‡,000*,000,000 * One million. This is a huge number. This, I'm guessing, is larger than most people can accurately conceive. ‡ Roughly the order of magnitude of the number of people on Earth right now. § Approximate US national debt, in $. ° The approximate age of the universe, in seconds. † Roughly the number of atoms in a gram of aluminum. ◊ Lower bound on the number of atoms in the universe. ¥ Upper bound on the number of atoms in the universe. ∞ Roughly the number of molecular events that have occurred in the universe, assuming each atom reacts once every femtosecond. There are essentially no physically relevant numbers above this. So... that number is about 10 billion billion *quadrillion* times larger than the largest number I would typically consider "even possibly physically relevant". The odds of the actual process underlying those pulse data are zero. All from a slight deviation and the extreme ends of the data.... Today I learned that when you crunch the numbers when estimating probabilities of hypotheses, it's really easy to get absolutely ludicrous probability ratios. Completely, utterly, bugnuts ridiculous probability ratios. 3) There's an effect sometimes called the "Monad tutorial fallacy", which comes from a problem in the functional language community that springs up around monads a lot. I should preface this by saying that I don't understand what a monad is. So if you're hoping to learn what a monad is from this TIL, I'm afraid you're going to have to wait (or go look up monads). Here's an apparently-common occurence when someone decides to learn about monads. The monad-expert-in-training goes and looks up an example of a monad. They struggle with that particular monad for a bit, then learn about another one, and then a couple more, and soon they've seen a lot of monads and they get an idea of what a monad is. Then they have an epiphany about what a monad *really* is, or what a really good metaphor is for a monad, and they get really excited. In their excitement, this person writes a tutorial explaining their metaphor -- for example, that monads are like burritos. This metaphor works perfectly well for the newly-minted-knight-of-monads, but anyone reading the tutorial gets *really confused* because they don't actually have any specific examples of what a monad might look like... they just know that monads are kind of like burritos. Except when they're not. Which isn't helpful. Thus, a new, confusing tutorial on monads is born. The moral of the story? It's easier to learn by abstracting from concrete examples than it is to try to learn the abstract idea first. Related: "Monads are hard because there are so many bad monad tutorials getting in the way of finally finding Wadler’s nice paper." -- Luke Gorrie (a hacker and entrepreneur) Also, it occurs to me that this series of TILs is exactly the kind of place an excited new monad-user would post a totally confusing summary of their epipheny on monads. If you notice me doing something similar, please call me out and direct me to this post.

May 24, 2016 at 04:29PM

Today I Learned: 1) According to a kind-of-sketchy-looking site called scienceheroes.com, there are four humans credited with saving the lives of more than a billion people -- Fritz Haber and Carl Bosch for the invention and commercialization of the Haber process, respectively (~2.7 billion people); Karl Landsteiner for discovering blood groups (A, B, O, +/0) (~1 billion people); and Richard Lewisohn for discovering a way to keep blood from coagulating outside the body with sodium citrate, thereby making blood banks possible (also ~1 billion people). scienceheroes.com also has an extended list. They credit 20 scientists with saving over 100 million lives. Of those scientists, 13 of them saved lives through vaccinations of one form or another. 2) Baby carrots are just normal carrots that have been sculpted into little cylinders. They were invented by a frustrated farmer in the 1980s, and now make up about 70% of all carrot sales. 3) There's a long-standing philosophical problem called Molyneux's problem involving blind people, as philosophical problems so often do. A blind person can instantly identify, say, a sphere or a cube by touch alone. If a blind person were to gain the ability to see, would they be able to identify a sphere or a cube by *sight*, using their knowledge of the object's shape? Unfortunately, this is a rather hard problem to solve empirically... ...which is why it wasn't solved until 2010. An MIT professor by the name of Pawan Sinha put together a project to restore the sight of five adolescent-to-teenagers blinded by terrible congenital cataracts, which previously limited their sight to essentially "it's dark out" vs "it's light out". The kids had their vision restored, then were tested on Molyneux's problem after less than two days of recovery. They could easily *distinguish* a cube and a sphere, but could not *identify* which was which with better than chance accuracy. They didn't do much re-testing after that, but what re-testing they did do (at up to 5 months with three of the five) showed that they learned to distinguish spheres and cubes quite accurately (80-90%, admittedly still worse than the average sighted person).

May 24, 2016 at 12:23AM

Today I Learned: 1) There are fewer carbon atoms (atoms!!) in a bacteria than there are transistors in a modern computer processor. Think about what that implies about computational efficiency of living systems... and also what that implies about human silicon-manufacturing technology. 2) The Earth is losing hydrogen at a rate of about 3 kg/second. Where is it going? It's literally floating off into space. Hydrogen is light enough that Earth can't permanently keep it grounded at typical global temperatures. 3) ...about the straddling bus, a proposed public transportation technology to rival the subway. A straddling bus is a bus. A big bus. A bus so big that it fits *over and around* the road, with a tunnel in the middle that cars can pass through (or allowing it to run right over cars). The idea is that it functions similarly to a subway, but less expensively (because it doesn't require building and maintaining a tunnel) and is more cool-looking. A straddling bus system was proposed at a 2010 tech exhibition in Beijing, and a test project was mapped out. Beijing authorities never gave the proposal a go-ahead, on grounds that the technology wasn't mature enough yet. This year, it's been re-proposed, and there's hope that it might get built. Thanks to Mengsha Gong for finding this fact!

May 23, 2016 at 02:21AM

Today I Learned: 1) Flails aren't real, maybe. Probably. I should clarify. Military flails definitely exist, but somewhere between most and all of them are artistic re-creations built far after the time period in which they were supposedly used. There are also contemporary-to-the-period artistic renditions of flails in use... but the same artists that drew the flails also drew some ridiculous, obviously-fantasy equipment in similar contexts, so it's really hard to tell whether they were meant to be taken seriously. More details here: http://ift.tt/1qhbLmH. This is one rare article where it's also worth going through some of the top comments. My favorite bit: "I suspect the flail had religious overtones and was therefore more of a ceremonial or "for show" piece for the pious knight." -> "Maybe like a weaponized censer? A 400 year old precuser to the over the top religious images of Warhammer 40K perhaps?". Thanks to Sarah Seid for teaching me a little bit more about medieval weaponry! 2) There's a new sensor for ATP! It's called "QUEEN", and it's similar to GCaMP, if you're familiar with that -- it's a modified verison of GFP (Green Fluorescent Protein) that's been split open, fused to an ATP-binding subunit of another protein (a bacterial ATP synthase, for the curious). When ATP binds to that subunit, it causes a conformational change and a stiffening in the subunit, which brings the two halves of the split GFP together and allows it to function. There are actually two versions of QUEEN, tuned for different concentrations ranges of ATP. QUEEN-7µ is linearly responsive between roughly 5 µM and 500 µM, and can be used to measure "low" concentrations of ATP (100 µM is pretty high concentration! Someone correct me if my math is wrong, but I think that's roughly an (average) inter-molecule spacing of about 25 nanometers.). QUEEN-2m is responsive between 1 or 2 mM and 10 mM, which is roughly physiological concentrations. Details behind the (sigh) paywall here: http://ift.tt/1TpHks2 3) ...about Bohmian mechanics, an interpretation of quantum mechanics in competition with the Copenhagen Interpetation (though never as popular). Bohmian, first developed by de Broglie and later championed by one David Bohm, postulates that particles are real, have actual, defined positions that don't require measurement to exist, and act non-locally. The theory trades the "weirdness" of wave/particles without defined positions and momentums for the "weirdness" of non-locality, i.e. explicit spooky action at a distance. For most physicists, the former bullet is easier to bite than the latter, and a paper in 1992 claimed to show that Bohmian mechanics predicted some nonsensical predictions of possible particle paths. A recent paper claims to show that the 1992 paper (known as the "ESSW" paper) didn't use Bohmian mechanics correctly, and that the "surrealistic trajectories" predicted by that paper actually make perfect sense if you account for the theory's nonlocality. Like the Copenhagen Interpretation (and like Many Worlds), Bohmian mechanics postulates a wavefunction that evolves according to the Schrödinger equation and (unlike Many Worlds predicts) collapses under observation. However, Bohmian mechanics doesn't claim that the wavefunction *is* particles the way other theories do -- instead, particles are real, have defined, precise positions and momentums, and interact non-locally. The evolution of particles follows a so-called "pilot equation", which is a deterministic function of the wavefunction and the positions of all the particles in the universe. I feel this is the time and place for a brief overview of the various interpretations of QM, according to Sam. Note that (as far as I know), all of these interpretations make the same predictions, for the simple reason that any theory that wasn't in experimental accordance with any other interpretation would be immediately discarded as wrong, since QM as a whole is one of the most experimentally well-supported theories in physics. The differences between interpretations is one of, well, interpretation, and what the theory implies about how reality is structured (spoilers: none of them are intuitive). The Copenhagen Interpretation says that the wavefunction is a full description of an unobserved system. When a system is measured by an experimental aparatus or observed by an observer, it "collapses" probabilistically to a point function (in violation of the Schrödinger equation, mind you!), essentially "picking" one state of particles in defined positions from the wavefunction, where the probability of picking any particular state is proportional to the (square of the magnitude of) the value of the wavefunction at that state. As far as I can tell, this is an essentially supernatural, dualist interpretation of QM. The Copenhagen Interpretation is still quite popular among quantum physicists, though nowhere near as popular as it used to be. Many Worlds (my favorite interpretation simply because it is the one I understand best except the Copenhagen Interpretation, which I'm fairly sure is nonsense) takes the wavefunction at face value -- it rejects collapse and instead proposes that all of the states of the wavefunction exist simultaneously. Many Worlds trades the ridiculous notions of "observers" and "collapse of the wavefunction" for the ridiculous notion that there are infinitely many parallel versions of the world playing out in accordance to the wavefunction. This is too much for some scientists, who claim that Many Worlds vilates Occam's Razor by postulating tons of extra worlds. Personally, I would counter that Many Worlds is a *strictly simpler* theory than Copenhagen and any other collapse-based theory, it just doesn't pretend to not notice all the impliciations of the wavefunction... but I digress. Bohmian mechanics proposes a wavefunction much like the Copenhagen Interpretation, but claims that the wavefunction guides the real, deterministic, particle-like evolution of *acutal* stuff, albeit in a nonlocal way. Bohmian mechanics trades the wave/particle duality for non-locality. Consistent Histories claims... something about how particles can have many alternate histories, and experiments give different weights to these histories...? Something about lots of matrix math and tensors...? I would really like to understand consistent histories, but I don't. Ensemble interpretation basically states that QM only applies to statistical ensembles of particles. Not sure what else it says. Relational QM, transactional QM (Object-oriented QM, anyone?), stochastic mechanics, objective collapse, many minds, quantum logic, quantum information theory, modal interpretations... no clue what these are. That concludes today's mini-rant on QM. Expect another in about six months.

May 22, 2016 at 02:25AM

Today I Learned: 1) The lac repressor protein, one of the best-studied and most widely-used repressor proteins in biology, operates at *extremely* low copy number, averaging around 10 proteins per bacteria. 2) We now know the details of photosynthesis in somewhat shocking detail, but have you ever wondered how we first learned things like the ratios of the different molecules used and produced in photosynthesis? The first such experiments were performed by putting leaves under water in specially-designed apparatuses that could capture gasses produced and measure their volumes/mass quite accurately. By adding different substances to the water, 18th century scientists could figure out what chemicals were required for photosynthesis (which released oxygen, which they could measure). 3) Star Trek fact: If you're familiar with Star Trek, you've probably seen the Klingon Bird-of-Prey. A lot. It's been the go-to Klingon ship design ever since Star Trek III. Today I learned that the Bird-of-Prey was concepted behind-the-scenes as a stolen *Romulan* vessel by the creators of Star Trek III, and it was assumed to be Romulan for some time. Why? I think because they wanted a ship that didn't look like Klingon ships from the TV series or the first movie, so they decided to make it not Klingon. You know, because each species in the galaxy has one ship. Except humans. Because we're special.

May 21, 2016 at 06:23AM

Today I Learned: 1) Not all gel-staining dyes go in the gel... some are meant to be added after the gel is run, by soaking the gel in the dye. 2) Solid chocolate has at least five different crystal types! The different crystal types represent chocolate in different molecular arrangements. Much like metals, which have different temperments depending on how they're heated, worked, and cooled, chocolate can take on different crystal structures, giving it different properties, depending on how it is handled. Type I and type II chocolate crystals are the highest-energy (least stable) chocolate forms. They are very soft and melt easily. That's not good for most chocolate applications, but it *is* good on cold things like ice cream, as it will melt more easily. If you eat ice cream dipped in chocolate, you probably want it dipped in type I or II chocolate. Types III and IV chocolate are moderate consistency and moderate melting temperature. These types are kind of crumbly, and don't produce a very satisfying snap, but they're reasonable for most applications and are the easiest forms to make -- they aren't as low-energy as type V chocolate, but if you cool chocolate under most conditions it will be kinetically trapped in type III or IV. If you leave type III or IV chocolate for long enough, it will slowly convert to the more stable type V, but this process excludes fats and some other molecules (type V effectively is less soluble to fats), which is why some chocolates will "sweat" if stored long enough. Type V chocolate is, most often, what chocolate makers aim for. Type V is very stable, and thus breaks with a highly satisfying snap. It also has a fantastic melting temperature of about 93°F, which means it will only really melt when you eat it or if it's hotter outside than you are inside. The melting process also absorbs heat, which supposedly gives type V chocolate a cool sensation. I'll be on the lookout for that effect next time I eat some good dark chocolate. The only real downside to type V chocolate is that it's hard to make. The standard way to make type V chocolate is to seed a molten chocolate mix with crystals of type V, which nucleates the production of more type V and, if done correctly, outcompetes the production of types III and IV. 3) Still on the topic of chocolate, did you know that coca pods are fermented before being processed into coca powder? Specifically, they are husked, piled up on the ground, and left to rot for a while. Fermenting chocolate is a bit like fermenting wine -- the exact conditions of fermentation determine the flavor of the chocolate, and chocolate made with specific techniques from specific places is highly valued among chocolate manufacturers.

May 20, 2016 at 01:27AM

Tody I Learned: 1) A couple of new fruit fly gene names: Htl (heartless) and Trol (terribly reduced optic lobes). Just a reminder, fruit fly genes are named after the phenotype that happens when you knock out that gene. Sadly, I find no evidence of a "moderately reduced optic lobes" gene, nor a "slightly reduced optic lobes" 2) You can buy pure E. coli ribosomes from New England Biolabs. Also RNA polymerase complex, with or without sigma-70. 3) About the "People Are Awesome" series of Youtube videos. Remember, all it takes to do something awesome is a very strong desire to do it and a willingness to try a lot of things to make it work.

May 19, 2016 at 03:15AM

Today I Learned: 1) ...a bunch of little things about FiveThirtyEight's coverage of Donald Trump, what they got wrong, why, and how they're likely to use that information in the future. I highly recommend reading this article -- it's a good window into the world of a data scientist and how data journalism works (and how it should work). Link here (not behind a paywall for once): http://ift.tt/1WDMvFJ Thanks Asher Rubin! 2) Eukaryotes seem to do a lot more signal multiplexing than prokaryotes. Bacteria tend to have one gene that detects a specific signal, which then turns on some downstream signal amplifier that triggers a cellular response. Eukaryotes have similar systems, but the responses of their detectors respond in much more complex ways, sometimes responding differently to different trigger signals. For instance, a stress response gene might start pulsing in response to ethanol, but turn on and stay highly-expressed in response to low glucose levels. This is a massive simplification, of course, but it's a mostly true one. 3) Bill Nye used to be anti-GMO, and anti-industrialized-farming. I say "used to" because about a year ago, he visited one of Monsanto's laboratories, talked with the scientists there... and now he supports the use of GMO crops. Interesting.

May 18, 2016 at 05:21AM

Tody I Learned: 1) There are a few password dictionaries out there for free download. You can find, for instance, the Rockyou-leaked passwords (among others) here: http://ift.tt/1cKEYyW. These are mostly (all?) illegally-acquired passwords that were put on the market, which the owner of the above link has cleaned of identifiable information and posted for free, for security testing purposes. You might want to try downloading the Rockyou password list and see if any of your passwords are in there. You might be surprised. Also, it's worth taking a look to get a sense of what kind of passwords people use. 2) A good rule of thumb to remember for Gillespie simulations (a kind of stochastic chemical reaction simulation based on molecular reactions) -- the more molecules you're simulating, the longer it will take. This is pretty obvious when you write it like that, but it really matters when the system you're simulating changes concentrations in the middle of a run..... 3) A species of Dictostelium (MY FAVORITE CLADE) called Dictostelium discoideum kills hostile bacteria by casting out a net of DNA covered with antibacterial agents when it detects a bacteria around. More details at http://ift.tt/27yu7RT (Nature paywall warning. =( )

May 17, 2016 at 04:02AM

Today I learned: 1) So there are agreements that DNA synthesis companies have to screen any DNA in orders they receive to make sure they aren't being asked to order dangerous stuff like toxin genes or genes for human pathogens. Today I learned that guidelines for screening are ONLY RECOMMENDATIONS, that all such screening is voluntary, and that the official recommendations for screening only apply to nucleotides longer than 200 base pairs. That means that if you can construct, say, a polio virus genome from only 100 bp oligos (which *I* certainly can), then you can do it with ordered DNA. On the one hand, that's a little dismaying. On the other hand, I think if someone's willing to go through the trouble of assembling a genome from 100 bp oligos, then simple screening processes aren't going to stop them. The screens are to stop amateurs and terrorists without a ton of biological expertise from simply ordering a deadly genome custom-built for them, without hampering the massively overwhelming majority of DNA synthesis that is totally harmless (and sometimes useful) 2) DNA has a theoretical storage density of 500 exabytes per cubic millimeter. For some examples of storage density and total readable capacity, see this cool figure: http://ift.tt/1Oxhw5V Here, "This work" is "Next-Generation Digital Information Storage in DNA" (http://ift.tt/1re3BNa if you have Science access, possibly http://ift.tt/1Aixs81 if you don't), in which the authors stored and read a book encoded out in HTML, images and all. They got 22 errors out of something like a megabyte of total stored information. Not particularly close to silicon-level accuracy by any means, but not bad for an in-vitro biological system. They used an encoding scheme where each nucleotide carried 1 bit of information (simply A or C for 0, T or G for 1 (or something like that)). The advantage of that system is that there's some flexibility in encoding, which is particularly nice because it lets you avoid long repeats of the same nucleotides, which are hard to both synthesize and sequence. There are other schemes that are a bit more compressed that can explicitly avoid repeats of any kind, which should make them even more amenable to this kind of storage. Why store digital information in DNA, you might ask? Reading off a hard drive is a hell of a lot faster and cheaper than synthesizing a bunch of DNA and then sequencing it. Hell, reading off a FLOPPY disk is faster and cheaper than syntheziging a bunch of DNA and then sequencing it, even if you don't own a floppy drive! Well, that's true, and DNA sequencing is unlikely to ever get faster than in-silico information retreival. BUT! The costs of DNA synthesis and information readout are plummeting much, much faster than the costs of in-silico data storage devices, per unit of information. In something like a few decades, DNA synthesis and sequencing should be *cheaper* than whatever the future equivalent of hard disks will be. Furthermore, DNA is actually really, really stable over long timescales, even relative to pretty stable stuff like hard drives. If you want to store information for a thousand years, DNA is a surprisingly good choice. Also, as the authors of the above paper point out, "DNA’s essential biological role provides access to natural reading and writing enzymes and ensures that DNA will remain a readable standard for the foreseeable future" -- in other words, your hard floppy drive might be obsolete and therefore effectively unreadable in 100 years, but you can be sure we're going to have technology for sequencing DNA as long as we have a civilization. So, in summary, DNA-based data storage is SLOW and expensive, but it will probably eventually be cheaper than hardware-based data storage, and it stores really well. It's a poor choice of storage media for your laptop. If, on the other hand, you want to take a snapshot of the internet and archive it for future generations, then your best bet might be to encrypt it into nucleotide sequences, synthesize the sequences, and bury a few tubes of the resulting libraries in some nice safe salt mines. 3) Windows has a cool tool for recording steps for reproducing bugs and errors. It's called the Problem Steps Recorder, and you can run it by going to the start menu and typing "psr". Just hit "record", do some stuff, and eventually hit "stop recording". The output is a weird pseudo-HTML file (zipped up, of course) that shows exactly what you did while recording, step by step, with pictures and automatic highlighting on the most relevant objects on the screen. The downside -- the weird pseudo-HTML can only be interpreted by IE, as far as I can tell, and IE can't even export it as normal, readable HTML. You were so close to impressing me, Microsoft. So close!

May 16, 2016 at 01:42AM

Today I learned: 1) ...about MOVE, a radical black liberation, anti-technology, anti-government, animal-rights-activist organization in Philadelphia in the 70s and 80s. MOVE consisted of one or two dozen members and their children, living as a commune in an urban, middle-class part of Phillie. From what I gather, they were somewhat extreme even as radical movements go -- they were well known for shouting obscenity-laced speeches at random hours on an amplified bullhorn, and their house was definitely armed and fortified. MOVE clashed with police in 1978, when the court issued an order to evict them from their house (I don't know under what charge). After a tense year, police officers tried to force their way into the house. An exchange of fire erupted and an officer was shot in the neck and killed. MOVE claimed that the officer was facing the wrong direction to have been shot by them, but the real point is that live fire was exchanged and multiple people were injured on both sides. The group moved after that, but the government didn't like them being in their new house any better (and the bullhorn-amplified speeches continued). They were eventually charged with contempt of court, violation of parole (presumably related to the earlier shootings), illegal posession of firearms, and making terrorist threats. In 1985, the Phillie police attempted to evict MOVE, bringing deluge hoses (courtesy of the fire department), tear gas, machine guns, and at least one anti-tank weapon. MOVE resisted coming out, so the police fired a few thousand rounds into the building. That didn't work, so they fire-bombed the house. With military-grade bombs. I'm not making this up. Most of MOVE (including most of the children in the house) died in the fire. Here's the extra-controversial bit -- remember those firefighters? The ones that brought deluge hoses? They were right there, and they were told by police not to put out the fire. As a result, 65 surrounding houses were also burned down. The official story is that the police didn't want to put firefighters at risk of getting shot, as there was still gunfire coming out of the house. The unofficial story was that they wanted to make sure everyone inside died, or something like that. The one surviving MOVE member was jailed on riot and conspiracy charges, though a federal jury later forced the city to pay her and her relatives $1.5 million in damages. The mayor later made an official apology for the incident. No officials were criminally charged. 2) Sumo wrestling has a fair number of spiritual rituals associated with it, most notably throwing salt into the ring and stamping feet to scare off demons. 3) T4 DNA ligase is super expensive!!! It's something like $5/reaction, which is way more than I expected from such a commonly-used enzyme.

May 15, 2016 at 04:04AM

Yesterday I Learned: 1) How to make Chipotle's sofritas! More or less. Actually, they were *even better* than Chipotle's sofritas, probably because of the effort and investment that went into them (and the extra lime (and maybe the triple dose of garlic)). Basically, the key is lots and lots of ancho chile sauce and cumin powder. Thanks Erik Jue, Anders Knight, and Andrey Shur for putting those things together with me! 2) It's quite possible that the harvester ants I'm keeping have one of the more painful known (and catalogued) stings, according to the Schmidt pain index. Fortunately, I have been able to avoid finding out for sure with some basic caution. Henceforth, I will use some extra caution as well. 3) You can buy wingless fruit flies online, and they're *really cheap*.

May 14, 2016 at 01:47AM

Yesterday I Learned: 1) Old thermal cyclers are huge because they're basically little refrigerators! Refrigeration means a pump, and some kind of effective form of heat dissipation. That means size. (Also, in case you didn't know, modern thermal cyclers use Peltier coolers, which are AMAZING devices that transfer heat the way a refrigerator does, but does it by applying electric current through a relatively small metal block. I honestly don't understand how the physics works, but the upshot is that they're small, light, and have no moving parts. The disadvantages are that they're expensive and not very power efficient.) 2) There's a eukaryote without any trace of mitochondria! It's called Monocercomonoides sp. PA203, and it's a single-celled, flagellated parasite, and by all rights, it really ought to have mitochondria. But it doesn't. Firstly, it's interesting how we know they don't have mitochondria -- we don't think they have mitochondria because they don't have any of the usual genes associated with them. That's a kind of odd way of telling if there are mitochondria... but it also is probably the most definitive way of doing so. If you look at them under a microscope, it's really hard to tell mitochondria apart from other organelles (and if you use a mitochondria-specific dye to mark the mitochondria and you don't see anything, it's hard to know if they actually don't have mitochondria or if they just don't stain the usual way). How does Monocercomonoides survive without mitochondria? Well, firstly, it doesn't seem to breathe oxygen -- all of the usual genes for aerobic respiration are missing, and it has an expanded set of anaerobic metabolic enzymes. That's the big thing that makes mitochondria necessary, but not the only thing -- apparently mitochondria are also critical for forming the iron-sulfur complexes used in iron-sulfur proteins... which are mostly used for aerobic respiration anyway, but apparently are necessary even without a mitochondria? Anyway, Monocercomonoides gets by using a set of four *bacterial* iron-sulfur complex formation genes. This isn't the first time this particular set of genes has been seen in a eukaryote, and the genes have introns in them, so they're not just bacterial contamination. Some of you out there have heard me claim that there really isn't gene transfer between bacteria and eukaryotes -- this is where I eat my shoe. Another important point the authors make is that this is likely the result of a *loss* of mitochondria at some point in the organism's past (which is awesome and strange) rather than a eukaryote that never had mitochondria in the first place (which would immediately make it one of the most important evolutionary findings of the modern era). How do they know? Well, they don't, but it's HIGHLY suggestive that a) every single one of this species' close relatives has mitochondria and b) all of those species (including PA203 itself) are parasites of animals. It's not particularly plausible that the only living eukaryote descended from pre-mitochondrial eukaryotes happened to evolve animal parasitism in parallel with all of its once-close relatives.... One question I have -- can these critters apoptose? As single-celled parasites, they may not need to, but I'm curious to know if they've evolved another way to do it. Now, bear in mind that all of the above comes from a hot-off-the-press Cell article, which in and of itself means it's more likely to be awesome (check) and more likely to be falsified in the next year (time will tell) than your average publication-in-a-decent-journal. So we'll see. Thanks to Asher Rubin for alerting me to this story! 3) Circadian rythems are more strongly affected by external purturbations at some points in their cycles than others. Unfortunately, the only data I've seen on this so far is from a cyanobacteria, so I have no idea how to apply this to human sleep cycles. (Another good fact to know -- most circadian rythems function normally for about 24 hours without any external signal before they fall apart.)

May 13, 2016 at 03:00AM

Yesterday I Learned: 1) The Echo Liquid Handler is a ridiculous machine. It's sort of a pipetting robot... except that instead of moving liquids around by pulling them up with a pipette tip and dispensing them, it works by spitting tiny droplets of liquid through the air with ultrasound. The machine uses two plates -- a source plate and a destination plate. The source plate is placed in the machine, and has wells filled with whatever liquids you're interested in moving around. The destination plate goes in *upside down*. The Echo moves the destination plate around so that a destination well is over a source well, then pulses the source plate with ultrasound to send tiny (order of magnitude 1-50 *nano*liters), precisely measured droplets up into the destination plate. Obviously, this limits it to working with pretty small volumes, but it's really fast (a few minutes to fill a plate), flexible, absurdly precise (supposedly), and works great at low volumes. The Echo also has one of the funniest failure modes I've ever heard of from a lab robot. The Echo can't handle bubbles on the surface of the source plate wells. The problem is that if it sees a bubble, it will interpret the *bottom* of the bubble as the top of the source liquid, and the *top* of the bubble as the bottom of the destination liquid, and will very precisely spit drops from the source well to the top of the bubble. 2) Relatedly, most plastics that are acoustically transparent (that is, transparent to ultrasound) are also visibly clear. Who knew? 3) DMSO has a very high surface tension, but low viscosity.

May 12, 2016 at 12:30AM

Yesterday I Learned: 1) Apparently radioactive animals are a serious problem for radiation containment. Rabbits contaminated by radiation, for example, have been known to leave cesium-filled poop all over the place, which is potentially a hazard... although if the rabbit can literally EAT that much radioactive material and still hop around long enough to leave that radiation elsewhere, I kind of wonder just how hazardous it really is.... 2) The gene Delta, like many genes involved in development, is named after the phenotype you see in flies when you knock it out -- in Delta's case, the knockout phenotype is wing veins patterned like a river delta. Given how crucial Delta is in a number of different cell types, all I have to say is... that's it?!?!?! (I guess I should expect the best-known developmental genes to have abnormally gentle knock-out phenotypes, as the ones with stronger knock-out phenotypes are harder to breed, therefore harder to study, therefore *not* the ones first characterized.) 3) Best estimates by climate change models predict that if the Earth warms by 4 °C, pretty much all of South America will become uninhabitable desert. As will much of the US. And China. And parts of Europe. And Africa. Global warming *could* cause reforestation... in the Sahel. The most barely-habitable of all currently-habitated deserts. How ironic.

May 11, 2016 at 03:02AM

Yesterday I Learned: 1) Commercial PCR buffers use dNTP (nucleotide) concentrations about an order of magnitude lower than what is generally found in vivo. 2) RNAs have a typical half-life of a few minutes in cell-free extract. Compare with proteins, which tend to not really degrade at all over a 1-day period, unless they're specifically tagged for degradation and special proteases are added to the mix. 3) Bacillis subtilis (B. sub) is a social bacteria with an interesting strategy for forming colonies. It has two modes of life -- a free-swimming mode in which it will move around and try to find good places to live sloughing off daughter cells as it does so, and a chained mode where it will stick to surfaces and grow long chains. By probabilistically switching between the two states, B. sub can plant lots of little colonies all over the place, giving it good odds of laying down a successful colony *somewhere*. Specifically, B. sub spends most of its time in the swimming mode, and will stochastically (and memorilessnessly) switch to chained mode. The transition from swimming to chained is totally random, but the switch *back* is actually pretty well-defined at around 7-8 cell generations after onset of chaining. That's important to keep the size of the colony relatively stable -- because the colonies grow exponentially, linear changes in *time* spent in chained mode result in exponential changes in colony *size*. Really tiny colonies of one or two cells are a waste of switching, and massive colonies of millions of bacteria aren't efficient either (some of them should be off exploring new potential colony sites). The well-defined chained-mode time keeps colonies at just about the right level, which is apparently a few hundred bacteria. Thanks to Andy Halleran for teaching me about this fascinating little slice of B. sub behavior! If you're interested in learning more and have access to Nature, check out http://ift.tt/1TEayx1.

May 10, 2016 at 04:34PM

Yesterday I Learned: 1) Bacillus subtilis (a commonly-used lab bacteria) has an interesting stress response. When B. subtilis (or "B. sub", as it's affectionately called in the lab) is stressed, it can respond in two different ways. B. sub can sporulate, meaning it encysts itself in some sort of hard, dehydration-proof case and goes into stasis, like a tardigrade. B. sub spores can last out whatever harsh conditions triggered the stress, only popping open when conditions are good again (apparently sporulated B. sub commonly survive being cooked in bread, and can cause the bread to go "ropy" (don't worry, though, B. sub is harmless unless you're severely immunocompromised)). Sometimes, though, a stressed B. sub can instead enter a state called "competency". A "competent" bacteria is one that is totally, utterly desperate for a solution to its problems, and tries to fix them by ingesting any nearby DNA in the hopes that some of it will contain a useful fix. Remarkably, there's enough DNA floating around, of enough variety, that this strategy is actually worth pursuing some of the time, and when it works, it means the bacteria can keep on growing -- much better than going dormant for years and hoping to land in a good spot to grow. It's still riskier than sporulation, though, and most of the time a competent bacteria is just going to die. The correct strategy for B. sub appears to be for most bacteria in a population to sporulate while a few enter competency and hope to get lucky. Yesterday I learned that B. sub has a more interesting competency pattern than I thought. It turns out that when B. sub becomes competent, it is virtually always a temporary state -- after a few hours, it swings back into sporulation and goes dormant. The genetic regulatory network that controls this behavior can be boiled down to a relatively simple model. There's a competency-triggering gene which, when high, triggers competency, and when low, allows sporulation (if other triggers are present to enter competency). The control gene activates the expression of two transcription factors that feed back to control its own expression -- a fast-acting activator and a slow-acting repressor. Normally, the competency gene is held at very low levels, where it fluctuates a lot. Occasionally, stochastically, it will happen to get produced enough to turn on its activator, which starts a feedback loop that kicks production WAY up and sending the cell into competency. Eventually, though, the slow-acting repressor kicks in and turns down expression of the competency-control gene. The cell goes back out of competency, and it can sporulate. The end behavior of this whole system is that B. sub can, if stressed, probabilistically and temporarily enter competency for a while before reliably coming back to sporulation, which is thought to be the (more or less) optimal strategy for survival under stress. 2) Matplotlib (python's most popular plotting module) has a pretty rich palette of built-in colors. There's a table of available colors in this stack overflow answer: http://ift.tt/21TVYIi 3) ...more than I think I wanted to know about printing labels....

May 09, 2016 at 12:36AM

Today I Learned: 1) Cas9 binds really, really tightly. So tightly, in fact, that it apparently doesn't let go even after it's cut its target! 2) While I'm on the topic of Cas9, Cas9 can also cut RNA... but it can't recognize PAMs made out of RNA, so you have to supply a little DNA oligo to anneal to the RNA's PAM, which Cas9 can hook to to start the binding process. Today I learned that this strategy works in vivo as well as in vitro. 3) Speaking of science things, did you know that canadian scientists have been censored on a number of environmental topics for nine years now? I certainly didn't. Apparently Canada's government had, up until very recently, required Canadian scientists to request approval any time they wanted to talk to the media about a number of topics. The rules for acceptable material to discuss were never made explicit, requests for approval took a long time, and scientists were very often effectively embargoed from communicating with the public. You know, I respected the Canadian government, overall, but this little piece of information just made them lose all credibility in my mind. Some details here: http://ift.tt/21MPGdl

May 08, 2016 at 04:38AM

Tody I Learned: 1) Tips for scrambled tofu: 1) "scrambling" the tofu with a fork gives it a more egg-like texture than cutting it up; 2) Instead of dumping spices directly into the pan, you can suspend them in water and pour them on like a sauce. This helps the dryness I typically get with scrambled tofu. 3) Speaking of dryness, it *is* also possible to get them too wet. It seems to be hard to get the water balance just right. 2) Tips for reducing primer dimers in PCR: 1) Raise the Tm, making it less likely for the primers to bind each other; 2) reduce the concentration of primers, making it harder for them to find each other; 3) increase the concentration of template, tying up more primers in the desired amplification. 3) An advantage of traditional paint media over digital that I had not thought about -- while you're making a physical painting, random stuff like paint drips or splashes can inspire you in new directions. Digital media typically won't do anything you don't explicitly tell it to, so it's a bit less randomizing and therefore less creativity-inducing. Just a small point, but an interesting one for a non-painter like me.

May 07, 2016 at 04:26AM

Today I Learned: 1) ...about the cheese fly (Piophila casei, which I assumes means something like "pie-loving cheese-protein"). I'm basically just going to quote the entire wiki article on cheese flies, because it is short and oh so sweet. Cheese flies are also known as "skip flies" or "skipper flies" because their larva can leap several inches to safety when alarmed. Those larva are also an occasional cause of disease (they can survive the stomach and worm their way into the intestinal wall, apparently mistakenly thinking it's a food substrate... ew) and culination (their fermentation products makes an especially pungent, apparently-delectable, and illegal cheese called "casu marzu"). As far as I can tell, there is nothing particularly remarkable about the actual fly. Thanks for the nightmares, Tara Sullivan. 2) ...what port tastes like. It's pretty good, for a wine. It's interesting having the contrast between super-sugariness and strong ethanol. Not something I'd drink on a regular basis, but interesting. Thanks for the port, Anders Knight. 3) Old thermal cyclers are *really heavy*!!!

May 06, 2016 at 12:28AM

Tody I Learned: 1) If you leave ice shavings at -20 °C for a couple of hours, the shavings will fuse into a more or less solid mass. Not recommended (unless you need a big hunk of porous ice for some reason. Hmm....) 2) ...another recommended food for fruit flies in culture -- cheese. Why am I not surprised? 3) The correlation coefficient between RNA expression and protein levels, across a bunch of different tissues, is around 0.35-0.5. That's... way lower than I expected. Makes me wonder just how useful RNA-seq really is (I mean, not really, it's still bloody useful, but still -- half of variance unexplained by RNA levels? That's more than I would have guessed). More (sadly in Nature, paywall alert) here: http://ift.tt/1SSTxje

May 05, 2016 at 02:06AM

Today I Learned: 1) ...what's up with halogen bulbs. Halogen bulbs are distinct from normal incandescent bulbs by having halogen gasses (usually iodine, bromine, or some mix of the two) mixed in the otherwise inert air in the bulb. Halogens are pretty darned reactive gasses -- why have them in a bulb? The answer has to do with bulb failure. Normally what happens in an incandescent bulb is that the heat of the bulb slowly causes the tungsten filament to vaporize. This causes two problems: 1) tungsten gas will adhere to the side of the bulb, eventually building up a a dark film that makes the bulb less bright; 2) as the filament loses mass and gets thinner, it becomes more succeptible to snapping when the bulb is rapidly heated, as happens whenever the light is turned on. Halogens in the bulb combat tungsten evaporation by quickly reacting with tungsten gas to form a tungsten halide that apparently either doesn't deposit on glass or does so reversibly (so that the halogens strip tungsten off the glass). At high temperatures, the tungsten halide will actually deposit *back on the filament*, preventing pretty much any filament loss over time. This makes halogen bulbs last longer than their incandescent counterparts. Perhaps more importantly, halogen bulbs don't dim over time, which is why they're used in laboratory equipment where even and predictable lighting is important (also stage lighting, though I'm not sure why it's important there....). Best of all, nothing in the bulb is particularly hazardous. Iodine and bromine aren't particularly *nice*, but they're not terribly toxic and they're relatively easy to dispose of. The major downside to halogen bulbs is that bit about re-depositing of tungsten only happening *at high temperatures*. Halogen bulbs run much, much hotter than their incandescent counterparts, which makes them a moderate-to-major fire hazard. 2) Vegemite is basically flavored, low-salt LB broth (yeast extract). 3) ...why my RNA cleanups have been failing (specifically, not yielding much in the way of RNA) -- I was accidentally skipping a step in the extraction, so my RNA was just falling right through the column. Also, ethanol contamination is a big problem with the kit I'm using. I'll have to take steps to keep ethanol out in future attempts.

May 04, 2016 at 02:04AM

Today I Learned: 1) ...what fruit fly eggs look like. Also, I learned that a way to induce fruit flies to lay eggs is to sprinkle a little powdered yeast into their food. Apparently they like that. The other way I'd heard to induce egg-laying is with a splash of vinegar -- sounds like they like fermentation products, or they think their larva will. Thanks to Andrey Shur for the fly tips! 2) There is a long and ancient European story of a race of headless men, or Blemmyes, which were said to have no head, instead possessing faces in the middle of their chests. These stories go back at least as far as ancient Greece, and were repeated and "verified" as late as by Sir Walter Raleigh. Go do a google image search for Blemmyes if you want an idea what they were supposed to look like. Just in case you are ever tempted to believe something is true because a bunch of people do, remember the Blemmyes. People believe a lot of weird stuff. Be skeptical. Thanks to Lady Ŀady Jade Beacham for filling me in on this! 3) So, I've been experimenting recently with some canned curry pastes. The recommended usage is that you mix the can with about a cup of coconnut milk, heat it in a pan, and throw it on stir-fry. Today, I wanted some curry, but I didn't have any coconut milk. Turns out that curry is much stronger (and spicier) without the coconut milk. Would I do it again? Depends on my mood.

May 03, 2016 at 02:16AM

Today I Learned: 1) Our autoclave has a light that tells you if the water level is too low... and if it's on, the autoclave won't run until the light has been dealt with and manually reset. Doh! 2) This is very much secondhand, and I don't claim to be an expert in Chinese mythology by any stretch of the imagination, but today I learned of a common pattern in Chinese mythology. You know how in Greek mythology (and, come to think of it, some other Western mythology too...) there's this common pattern where Zeus or another god (but let's be real, it's usually Zeus) gets the hots for some mortal woman and decides to come down and impregnate her before taking off and leaving her with a semi-dietic child? In Chinese mythology, the pattern is more like this: spirit/god falls in love with a mortal. Spirit/god's compatriots disapprove of spirit/god's love. Spirit/god goes into hiding in the mortal world so he/she can be with his/her love. Spirit/god is found out and forcibly separated from his/her love. I feel like stories of god/human love were meant to serve *rather different purposes* in China and in the West.... Thanks to Mengsha Gong for teaching me a little folklore outside my culture! 3) A trick for cleaning up colony PCRs -- dilute the colony in ~20 uL of water before adding it to your reaction. I can't yet speak to the effectiveness of this tip, but it's a tip.

May 02, 2016 at 01:39AM

Today I Learned: 1) The word "bleb" is not in common usage in English. Apparently it almost exclusively applies to cells undergoing, uh, blebbing. 2) How to tie off the end of a thread. Badly. But hopefully well enough. 3) You can anesthetize fruit flies by sticking them in the freezer for a few minutes. Once they stop moving, you can keep them on ice for a while, and they'll recover quite quickly once heated to room temperature. Tons more information on culturing fruit flies at http://ift.tt/1SHGwcn.

May 01, 2016 at 03:28AM

Tody I Learned: 1) ...how to use Adobe Fireworks, or at least a few tricks with it. I now know how to use the magic wand tool, apply textures, make my own textures, move things around, and a few other things. Perhaps my favorite is rubber-stamping, which is essentially a brush that copies content from a location relative to the cursor (in a way you define). 2) The definition of "persistence length". Informally, persistence length is the distance over which a long, thin object remains stiff and rod-like. For example, the persistence length of rope is quite short -- only a few inches at most -- while the persistence length of spaghetti is approximately 10^18 meters. That means that at distances longer than 10^18 meters, spaghetti acts more like a string than like a rod. Who knew! Anyway, today I learned the exact definition of persistence lengh. It has to do with the average angle formed between two ends of a rod of material. As the material gets longer and longer, the end wobbles around more with respect to the beginning. Specifically, it can be shown (somehow, I don't know how) that the average cosine of the angle between the two ends falls off exponentially. The persistence length is the constant that defines the rate of that dropoff. More precisely, if A is the angle between ends and L is the length of the rod, then P is the constant such that mean(cos(A)) = e^(-L/P) This is a nice tidy definition mathematically, but it *does* make it a little hard to intuit exactly what it means. From another article (http://ift.tt/21n39YY), the persistence length is also the distance at which the average angle between ends reaches 68.4°. 3) The average mass of protein in a cell membrane can be higher than the average mass of lipid in the same membrane. Whoa. ...of course, this might be a little misleading, since membrane proteins generally have stuff sticking out of both ends of the membrane. Oh, another useful fact -- typical membrane thickness is about 4 nanometers, which means the trans-membrane domains of membrane proteins are also about 4 nm.