March 01, 2016 at 12:49AM

Today I Learned: 1) ...what chartreuse looks like. It's a slightly yellowish lime green. Technically, that's web chartruse, a.k.a. chartreuse Green, a.k.a. 7FFF00. There's also *chartreuse yellow*, a.k.a. "traditional charteuse", a.k.a. DDFFF00, which is a much yellower shade that's rather harder to define. 2) The two most common causes of unfair dice (at least, due to manufacturing error) are 1) two opposite sides being squished slightly together, causing those two sides to be more highly weighted, and 2) air bubbles trapped in the die plastic, which caus sides near the bubble to come up more often because they're lighter. 3) An AU (Astronomical Unit, the distance from the sun to the Earth) is, very roughly, 10^-5 lightyears.

February 29, 2016 at 12:08AM

February 28, 2016 at 02:38AM

Today I Learned: 1) The output of Niagra falls is mandated by law, which isn't as ridiculous as it sounds. Much of Niagra is diverted for various kinds of water use (I think mostly electricity), but there's a minimum amount of water that must be left flowing so that it's still, well, Niagra falls. That amount, by the way, is 100,000 cubic feet per second. I have no idea how much that is, except that it's about as much water as Niagra falls. 2) When given a choice between vegan "chicken", vegan "steak", and vegan tofu (er, just tofu I guess), I choose the tofu. I don't even think about it. The other two barely sound like food. 3) Solidworks tip -- anchor your design to an axis. Otherwise, your design could accidentally be underconstrained, so when you make changes, it might have unpredictable consequences.

February 27, 2016 at 12:39AM

Today I Learned: 1) RNAzol, a compound used for extracting RNA from cells and tissues, absorbs with a spectrum JUST SIMILAR enough to RNA to make one think it might just be RNA... don't be fooled by the ng/uL amounts, though, it's just contamination! 2) Speaking of RNAs, it seems that when preparing an RNA ladder for a gel, you should let it cool for more than 90 seconds on ice after melting it. I'm still not sure if it's important to use a long, cool melt. 3) EDTA is freaking hard to solubilize! I was setting up a protocol that called for 0.5 molar EDTA, which I didn't have. "That's fine", I thought to myself, "it's just a salt in water. I'll just whip together a solution in my 10 minute wait step". Thirty minutes later, I'm googling "dissolving EDTA" and reading all about the difficulties people have solubilizing it. Apparently it won't dissolve unless you raise the pH (usually with sodium hydroxide), which is hard because EDTA is a relatively strong acid. Even with the pH high enough, it usually takes heat, constant stirring, and patience. One of these days, I'll go ahead and make enough EDTA solution to last the lab until I'm done. Until then, I'll be either skipping EDTA or using TE buffer instead.

February 26, 2016 at 12:17AM

Today I Learned: 1) RNA ladders* are really oddly tricky to use correctly. I'd had some trouble getting mine to work before -- they were coming out as a smear, rather than as nice crisp bands. I figured the ladder must have been contaminated by RNases and ruined, so I went and got a fresh ladder, a fresh set of pipettes, and a whole bunch of RNAse-away. It didn't help. What *did* seem to help was to change the preparation protocol a bit. New England Biolabs, who sells this ladder, recommends heating it either at 90C for 2 minutes or 70C for 10 minutes. I had always previously done the former, but the ladder came out much more nicely when I did the latter. But then I tried it again later in the day and it didn't work... I think maybe I didn't let the second one cool long enough before loading it, but I'm really not sure. * A ladder, in the context of biology, is a standard set of DNA, protein, or, in this case, RNA of known size and composition that can be run alongside experimental samples in an electrophoretic gel. Since you know the sizes of the ladder components, you can use their bands on the gel to determine the sizes of your own samples. 2) There's a virus, called an iridovirus, which infects crickets, and causes them to mate much more frequently (until it kills them). It also happens to sterilize the crickets, presumably so the cricket spends less time laying eggs and more time mating. Oh, and it turns their guts blue. 3) Guess how scientists first discovered that some frogs are venemous? A scientist picked one up and it bit him. The last two facts courtesy of this entertaining NPR article: http://ift.tt/20YAs2S

February 25, 2016 at 12:26AM

Today I Learned: 1) It really can't hurt to ask. 2) A mark above or below a character to change its pronounciation (like the degree symbol in å or the "'" in á) is called a "diacritic". 3) Most standard systems for making guide RNA plasmids involve a linearized plasmid backbone with a promoter at one end (pointing off the end) and a tracrRNA sequence at the other end. To build the vector, you buy phosphorylated primers with your target sequence and just blunt-end ligate (or, probably, restriction digest/ligate, depending on the system) it onto the ends of the backbone, and voila! You have a plasmid that produces a hybrid guide/tracr RNA!

February 24, 2016 at 02:36AM

Today I Learned: 1) How many replication origins do you think there are in the human genome? That is, at how many points in the genome does the cellular replication machinery bind and start duplicating the chromosome? If you asked me over dinner, I would have guessed one per chromosome, so 23 (well, 46 under most circumstances, given diploidy and all). On further reflection, I might have answered "a few per chromosome", because if there are multiple origins, there can be multiple replication forks, which means the cell can duplicate its genome faster. The real answer? 80,000. I couldn't capitalize that for emphasis, since there are no capital numerals, so let me try that again -- EIGHTY THOUSAND, or about THREE AND A HALF THOUSAND per chromosome. Oh, and double that in a diploid, i.e. amost every human cell. That's a lot of replication forks. Incidentally, from this we can calculate the approximate deoxynucleotide consumption rate of a human cell undergoing replication. According to bionumbers (ID 104136), a single repilation fork adds about 33 nucleotides per second. This is pretty consistent with bionumbers ID 111200, which claims a replication fork velocity of about 1.5 kb/minute. With 80,000 of these forks running at once, and two strands synthesized per fork per base pair, that's a nucleotide consumption rate of about 5 million nucleotides per second per cell. If every cell in your body were to go into division at once (don't try this at home), that would be order of magnitude 5E6 * 4E13 ~ 2E20, or 200 quintillion, or 200,000,000,000,000,000,000 nucleotides per second, which is also about a third of a millimole of nucleotides every second. Nucleotides average about 487 grams per mole, so that's something like a milligram of nucleotides per second, which is about the weight (per second) of a medium-coarseness grain of sand (or a microliter, if you're familiar with those). 2) The US has the largest air force in the world, by a fair margin. No surprise there. Guess who has the second-largest air force in the world? No really, guess. ... ... It's the US navy. Thanks to Anders!!!! on this one. There's some discrepancies between different wiki pages on the subject of air power -- check out http://ift.tt/1hYFW8X, http://ift.tt/1Lb4O08, and http://ift.tt/1On2lQC -- so I'd check the numbers yourself to be sure. 3) Our country's (the US's) first war was fought alongside the French against the British government. Our *second* war, the Quasi-War (yeah), was fought *against* the French government, ostensibly on the side of the British but without much coordination or cooperation (they sold us some arms and we shared secret signals to keep from attacking each other). The Quasi-War was a roughly two-year naval war mostly consisting of massively effective pirating of US merchant vessels by the French and massively effective raiding, capturing, and sinking of French military vessels by the US navy. The war has its roots in the rather massive monetary debts owed to the French government by the US for loans paid during the Revolutionary war a couple decades before. When the french government was toppled in their own Revolution, the US (not sure whether it was ambassadors, legislators, or the president) decided that they weren't going to pay back their debt because it was owed to a previous government, not to the current revolutionary government (also, the US was trying to repair ties with Britain). This led to some rather antagonistic diplomatic exchanges between the two countries, eventually leading the French to decide they'd get their money back in the standard way that states seem to have extracted money from one another -- by privateering (state-sponsored piracy). You know, it seems like the US got into a lot of wars over its merchant fleet. The Quasi-War was the first. In the two years before a treaty was signed, the US fleet of about 25 naval vessels won a number of engagements against the french, losing only one ship (which was later recaptured). This, however, didn't stop the French from capturing over *2000* merchant ships. So really, the whole thing was a bit of a win-win scenario -- France got its money back, and the US navy gained some much-needed credibility in the European stage.

February 23, 2016 at 01:02AM

Today I Learned: 1) Flu kills about half a million people a year. Half a million! That's... surprisingly large. Also, fever is a really good marker of having the flu -- if you have a fever and a cough during flu season, odds are around 80% that you have the flu. 2) Spiegelman's Monster is GGGGGCCUUCCUUGCCAGACCGCCCCAAUACAGUAACAGCGCGUUGGC. It is the smallest known RNA sequence capable of replication with only an enzyme (specifically, the RNA polymerase of bacteriophage Qβ, which is unusual for replicating RNA from and RNA template). It was derived (by Spiegelman) from the sequence of Qβ, which is an RNA virus, by putting a bunch of Qβ RNA genome in a test tube with Qβ RNA polymerase for a while. The genome replicated a whole bunch, occasionally with mutations. Shorter variants replicate faster in such an environment, so there was selection for short mutants. In the 1965 paper, the shortest sequence the researchers got was 218 bases, after a mere 74 generations of replication (I can't give you the sequence because it wasn't sequenced -- this was 1965, after all). A later replication of the study wen longer and eventually got the sequence I showed above. This was one of the classic experiments of minimalist evolution. 3) Ribozymes (RNA molecules that function as some kind of enzyme) may be much more common than I thought in sequence space. A 1993 experiment by Bartel and Szostak tried discovering new RNA ligases from random 220-base-pair RNA sequences. They built a library of about a quadrillion (10^15) random RNA sequences, then selected for sequences that could catalyze a particular ligation reaction. After four rounds of in vitro selection (without mutation) they eventually isolated *68* distinct RNAs capable of ligating about 10,000 times faster than with no catalyst. That's not particularly good, for an enzyme (it's terrible) but it's quite good for an RNA, especially a random one. With error-prone PCR and a bunch more selection, they were able to get ligases with activities of several million times the background rate. Assuming the random RNAs were actually representative (which I'm actually pretty willing to grant in this one), that implies that about one in every 15 trillion (~10^13) RNAs is at least a semi-functional ligase. That's really impres sive. Given that there are 4^220 ~ 10^132 possible RNAs of length 220 nucleotides, that implies that there are somewhere around 10^119 RNA ligases somewhere in sequence space. That's a kind of number big enough to start comparing to the estimated number of "molecular events" that have happened in the entire unvierse over the entire history of the universe... and that's just ligases! Think of what else RNA can probably do!

February 22, 2016 at 01:32AM

Today I Learned: 1) Ok, so I actually learned this fact yesterday, but it's so good I want to cheat and share it today instead. It has to do with the aardvark (which, incidentally, has "least concern" conservation status and is doing quite well) and its very special symbiotic relationship with the aardvark cucumber. Yes, the ardvark cucumber. The aardvark cucumber is a plant that has evolved to *specifically be dug up and eaten by aardvarks*. Aardvark cucumbers (which are, in fact, a kind of cucumber) have really succulent, delicious, water-filled cucumber fruit that look kind of like puffball mushrooms. Unlike most cucumbers, the aardvark cucumber only grows its fruit underground, where it's inaccessible to most mammals with the exception of, you guessed it, the aardvark. Aardvarks LOVE aardvark cucumbers, apparently mostly because they're a really important source of water for the aardvark. This is unusual, because aardvarks don't really eat any other fruit -- aside from the aardvark cucumber, aardvarks pretty much exclusively eat ants and termites, and only certain varieties of ants and termites at that. So the aardvark cucumber must be something special. This arrangement of being dug up and eaten by aardvarks works out for the cucumber, too. Aardvark cucumber seeds germinate best when buried in freshly fertilized, loosely tilled soil. Aardvarks dig up the cucumbers, poop out the seeds nearby, and bury them, giving the cucumber exactly what it needs to germinate. The aardvark cucumber appears to have no other mechanism for spreading its seeds -- if the aardvark ever goes extinct, it will take out the aardvark cucumber with it. Aardvark aardvark aardvark aardvark aardvark. 2) When writing Firefly, Joss Whedon wanted to the cast to use Chinese as a second language, particularly during particularly outbursts of intense emotion, most of which ended up being cursing. The only problem was that that Chinese is super syllable-efficient, so Whedon had to come up with really long, involved curses for them to last long enough for an English audience to pick up on as Chinese. 3) Pay-for-performance is an idea in healthcare that you can get better performance from doctors by paying them more for better patient outcomes. It's an interesting idea, but so far empirical evidence says that it only works very slightly, if at all.

February 21, 2016 at 02:32AM

Today I Learned: 1) How to use Solidworks! ...ok, not really, I've had two tutorials in solidworks before, but today was my first time building something new in Solidworks from scratch, with no guidance. I think it's even what I intend! Will need Erik!!!! to check me on that, though. (Also, for those who don't know, Solidworks is a program for designing 2D and 3D models of things; for my purposes, it's a tool for drawing stuff to 3D print) 2) KL740 E. coli grown at 29 C on LB-ampicillin plates take somewhere between 12 and 36 hours to grow colonies, according to my most recent batch of transformations. At 12 hours there was no visible growth at all, and I thought they'd all died. At about 38 hours, there were big, fat, overgrown colonies. By 45 hours, there were little tiny satellite colonies. Now you know. 3) ...how to make methanol from CO2, enzymatically. You bind three enzymes (formate dehydrogenase to convert CO2 to formate ; formaldehyde dehydrogenase to convert formate to formaldehyde; and alcohol dehydrogenase to convert formaldehyde to methanol) into a sol-gel (basically an aerogel), then run a bunch of CO2 through the gel. You can also bind each of the enzymes onto a separate platform, as I've described in an earlier TIL, and do the conversions one at a time. Details on the former here (http://ift.tt/21hulZA) and the latter here (http://ift.tt/1TxBio6) (sadly, probably behind a paywall).

February 20, 2016 at 05:46AM

Today I Learned: 1) Ever seen pictures of the sun? Did they look like a perfectly round, featureless disc? If not, then you weren't seeing visible light pictures. Turns out you can figure out a lot about the sun by looking at it in different wavelengths, largely but not entirely because different layers of the sun are different temperatures and therefore emit and absorb different wavelengths. Almost every wavelength looks way cooler than visible light (except microwave, the microwave pictures are pretty featureless). Relatedly, the surface of the sun is the coldest part of the sun -- not only is it cooler than the interior, which isn't surprising, but it's also cooler than the plasma around it, which is surprising. It has something to do with how plasma escapes the sun. Turns out the sun is full of tangled up, high-energy magnetic lines. You could think of it like a million-kilometer-diameter ball of rubber bands, except those rubber bands are constantly roiling past each other and are thicker than our planet. These sometimes break the surface, spewing extremely high-energy plasma out into space -- that's what we call a solar flare. That's also what populates the "atmosphere", I guess you could call it, of the sun, which is VERY BIG and VERY HOT. It's best viewed with UV and X-ray telescopes, which is kind of absurd. Anyway, the point is that with different wavelengths of light, you can see different layers of the sun, down to just below the surface. There is also a technique to see what goes on *under* the surface, called helioseismology. Basically, all of that intense roiling I mentioned causes a lot of earthquake-like vibrations in the sun (http://ift.tt/1HLU2uh) which can be tracked by telescope. With some presumably very fancy math, you can track individual seismic events as they propagate through the outer third of the usn, just like you can track vibrations in the Earth when there's an earthquake. By tracking deflections of those vibrations, you can reconstruct what's going on under the surface, at least a third of the way in. Wait, you ask, why only a third of the way in? Well, apparently the outer third of the sun is the "convctive layer". Inside it's way hotter and the matter in there... doesn't convect vibrations. They *should*, in theory, be able to transmit *gravitational waves*, but that's currently impossible to detect because we can't see the inner two-thirds of the sun. Solar astronomy is awesome. 2) More Sol facts! The sun takes up approximately 1 arcsecond of the sky. Someone should check me on this, though, I haven't actually double-checked the math. 3) Another Sol fact! Light emitted by matter at the sun's core can't go to the surface. It will run into stuff and either get reflected or get absorbed and re-emitted (which I *think* are fundamentally the same thing, but my QM isn't strong enough to say that for sure). In this way, the light bounce around randomly, eventually random-walking its way to the surface. This process, from the core of the sun to the surface, takes order-of-magnitude 100,000 years. Yeah. Bonus fact) In astronomy, "metal" commonly means "anything heavier than helium". I guess when you're talking about stars, it really doesn't matter what chemistry an element has at STP.

February 19, 2016 at 03:08AM

Today I Learned: 1) Three-primer PCR. One is the normal forward primer. One is a normal reverse primer, with an extension. The third is a primer that amplifies off of the extension from the reverse primer (or it could be an extension off the forward primer). This is nice if you want to make two different kinds of modifications to the end of a sequence combinatorially. This came up today when discussion how to mutigenize near the end of a linear DNA sequence while also adding two or more different sets of overhangs to the same end. 2) I apparently had really good English teachers in high shcool. There was a discussion over dinner today about the many awful experiences my coleagues had in high school, and while I could sympathize, I can't really say I had those experiences, for the most part. All of my high school English/Lit teachers were pretty awesome! 3) You know how cheap bamboo chopsticks have that solid block at the end farthest from the eating end? Apparently that's not just to hold them together, and you're not supposed to split it when you break apart the chopsticks. Instead, you're supposed to break off that whole piece, and use it as a stand for your chopsticks so they don't have to touch the table.

February 18, 2016 at 01:16AM

Today I Learned: 1) Apparently there are at least some plasmids that can replicate even when host protein synthesis is inhibited with antibiotics. Source: Andrey!!!!! 2) In a closed system without aeration (say, a spacesuit in space), a human will die from CO2 poisoning well before he or she runs out of oxygen. Source: The Martian. 3) The active ingredient in roundup, glyphosate, has recently been declared a probable human carcinogen by the WHO. However, there's a HUGE amount of controversy over this particular chemical. There have been a bunch of epidemiological studies on roundup, which have suggested everything from "it's harmless" to "working with it occupationally increases risk of Hodgkin's lymphoma". There is no direct evidence of glyphosate's toxicity in humans, BUT there are animal studies that suggest that it may be. More details here: http://ift.tt/1GT5TD2. It is definitely NOT acutely toxic in any real sense -- the LD50 in rodents of various kinds is on the order of grams per kg, which is pretty close to that of 70% ethanol (i.e., if you drank a bottle of it, you'd be about as likely to die as to live). (also of note: the WHO has no water guidelines for glyphosate concentration in water because they claim the lowest known health-safe dose of glyphosate is orders of magnitude above anything commonly found in drinking water)

February 17, 2016 at 12:59AM

Today I Learned: 1) A good method of finding a plane to separate several corners of a cube (or other parallelopiped) from the other corners. In brief: * Draw out the cube, mark the corners you want separated. * Try to draw points in the edges that separate those corners. * Connect the points into a plane, if possible (having the edges marked makes this much easier) * Pick three of the intersection points and make a rough guess at their coordinates. This works best if your cube is arranged nicely with respect to the basis axes of your space. * Now that you have three points, you've defined a separating plane! Just remember that a plane satisfies aX + bY + cZ + d = 0, plug your three points into this, solve for the coefficients (you'll have one free parameter that you can set however you want), and you have your separating plane. 2) The Grant Study and the Glueck Study are probably the two longest-running social experiments in, uh, social science. The two parallel studies followed several hundred young, white, male Harvard undergrads (Grant Study) and disadvantaged Bostonian inner-city youths (Glueck Study) for 75 years, and as far as I know is still running. These studies have required an absolutely ridiculous amount of legwork and has been overseen by at least four separate directors. The main results can be found on the wiki page for the Grant Study, but it can be roughly summed up as: alcoholism is really bad, having good-quality relationships is really, really, really important. 3) Corynebacterium glutamicum is a popular bacteria in industrial production. It can produce a wide variety of compounds, and it's prarticularly hardy under anaerobic conditions with lots of metabolic byproducts (basically, they continue production in stationary phase). Just in January, Korean scientists engineered a strain of C. glutamicum capable of metabolizing xylan-based hemicellulose, which is a completely novel carbon source for that species. Amazingly enough, none of my facts came from Anders Knights today! Maybe next time, Anders.

February 16, 2016 at 02:26AM

Today I Learned: 1) Guinea worm is a terrible, awful disease, and it's probably going to be the second disease intentionally eradicated by humankind (after smallpox). Over the last thirty years, the number of cases of guinea worm has been brought down from about 3.5 megapeople per year to 22 per year. 22! The striking thing about the program to eradicate guinea worm is that it involves no new technology and essentially no treatments, because there are no treatments for guinea worm. The eradication program works pretty much entirely by convincing people to filter their drinking water and keeping infected persons away from other people's drinking water. And that works (though it's a lot of effort). Also, the person most singly responsible for wiping out guinea worm? Jimmy Carter, via the Carter Foundation, which leads a bunch of other organizations you've probably heard of in the campaign to wipe out guinea worm. Quote from Carter, while sick with cancer: "I would like the last guinea worm to die before I do". Hardcore, Jimmy, hardcore. 2) The Stokes(-Einstein) radius of a molecule is the radius that a perfectly spherical molecule would have to have to diffuse at the same rate as that molecule. Along with this definition comes an interesting revelation -- at least for practical purposes, even a molecule with a complex shape behaves as though it were a sphere of some radius for the purpose of diffusion. Yeah, Anders Knight gets this one. 3) It is SOMETIMES possible to boot a computer that fails an fsck due to a mismatch in partition sizes according to different sources by telling it to ignore the error. It feels pretty dirty, but it worked for us this time....

February 15, 2016 at 01:11AM

Today I Learned: 1) There's a class of fluorescent proteins called fluorescent timers, which are used to... time things. Sort of. The primary feature of a fluorescent timer is that it changes color over time. First-generation fluorescent timers were derivatives of mCherry (a red fluorescent protein, but not Red Fluorescent Protein (RFP)) that changed color from blue to red as they matured. These timers tended to clump, though, making them not that useful. More recent fluorescent timers (first made back around 2012, from what I can tell) are made from fusions of two fluorescent proteins -- for instance, a fast-maturing GFP fused to a slow-maturing mCherry. If you look at it and it's green, then it's a young protein. If you look at it and it's green AND red, it's old. By tagging various proteins with the fluorescent timer motif, you can do some pretty cool experiments. For example, it turns out that yeast cells preferentially move older copies of at least several different proteins into their daughter cells when they bud, which we know because you can tag those proteins with fluorescent timers, and you see more red/green ratio in newly-budded cells than in their parents. What would *you* like to do with a fluorescent timer? 2) You know that idea of ether that was disproven conclusively back in the... 20s? Nope, wow, I just learned that the Michelson-Morley experiment was performed in 1887! That's quite a bit earlier than I thought. Anyway, what I *meant* to write about was exactly what feature of the ether was disproven. See, as a canny poster on Reddit's AskScience noted, the idea of ether, an all-pervading medium in which light propagates, is oddly similar to the idea of, say, the electromagnetic field of quantum field theory, which is an all-pervading medium in which light propagates. The important thing that was disproven by the Michelson-Morley experiment was the idea that there might be a *universal reference frame* in which light moved. There is no such reference frame. 3) The Star Wars Customizable Card Game (SWCCG, for short) still has a tiny but active player base, and a team of community members that makes rulings, runs tournaments, and "prints" new cards. It currently has an actively maintained comprehensive rulebook that appears to rival that of Magic: The Gathering in size and complexity, a fact that would have been nice to know back when I played. Just so you understand what this means to me, the SWCCG was THE card game between several of my friends and I during elementary school. The game was made by Decipher Inc., which has always been good at making incredibly resonant games based on established franchises (they made a Star Trek card game that was, in its day, one of the most popular trading card games behind Pokemon, Magic, and SWCCG; they later made a Lord of the Rings card game that was mechanically fascinating but outcompeted HARD by the Magic juggernaut and several other popular games like Yu-Gi-Oh that were establishing themselves at the same time). SWCCG was an interesting mix of awesome-looking cards, innovative mechanics, and a slavish adherence to flavor that, while cool, resulted in some of the weirdest mechanics of any card game I've ever seen. For example, there were several cards (H'Nemthe is the one I remember) that specifically affected cards representing male characters. How do you know if a card is male? From the rulebook: "Only characters (even droids) have gender. To determine the gender of a character, examine title, lore, and game text for words which will indicate the gender (he, she, him, her, male, female, princess, etc.). If there are none, check the picture and see if a reasonable person would conclude that the character is female (if you are not a reasonable person, find one). If it is unclear, the character is considered male." Yeah. Also, there were a bunch of mechanical traits that could only be determined by reading the card's flavor text -- to find out if a character was a spy, you check the flavor text for the word "spy". I could write at some length about the idiosynchracies of the SWCCG's design, but suffice it to say that it was one of the coolest, most poorly-designed-but-surprisingly-fun-anyway games I've had the pleasure of playing. It was also among the most popular card games (#2 behind MtG for a while) before ridiculous power creep kind of killed sales and poor money management caused Decipher to more-or-less implode, leading to the game's official demise around 2002 or 2003. I'm glad somebody's keeping it alive.

GFP Folding, pcs (not PCs), and GFP Maturation

Today I Learned:

1) Apparently the time required for a genetic biocircuit to run to completion with a circuit using GFP or a similar protein as the output is mostly dominated by the time it takes for GFP to fold. A new paper from Vincent Noireaux includes, among many other things, a series of experiments expressing one- or two-layer gene circuits (with GFP as the output) in a cell-free transcription/translation (TX-TL) extract. The one-layer circuit was just GFP, expressed under a strong promoter. The two-layer circuits were variations of "GFP under an activatable promoter and the activator under a strong promoter".

The striking thing about their results, to me, is that moving from one layer to two layers results in no more than a 50% increase in time required to run, looking at it by eye. Check out figures 8a-8i in http://ift.tt/1RyeKlC to see for yourself.

2) The standard abbreviation for a parsec is "pc". I was also a little fuzzy on *exactly* what a parsec was, so I checked in on that again -- a parsec is the distance that the end of a pole 1 AU* would make if you pivoted it through 1 arcsecond**. EDIT: Nope, this is wrong -- it's the *length of a pole* required so that when you swing it one arc second, it traces out a distance of 1 AU. Also known as a few lightyears.

* ...and an AU is an Astronomical Unit, or the distance from the Earth to the Sun, or about 8 light-minutes, or 150 gigameters, or 150 million kilometers.

** ...and an arcsecond is 1/60th of a degree, or 4 microradians, or the angle spanning about 2 kilometers on the Earth's surface as measured from the Earth's center, or about 4 seconds' worth of time on a clock face.

3) Above I stated that GFP (green fluorescent protein, the standard fluorescent protein) takes a long time to fold after being translated. Turns out that's not really true -- GFP folds relatively quickly. It *does*, however, take a long time to *mature*, which is not the same thing. See, GFP does something really interesting to become fluorescent. The fluorophore in GFP is actually just a single modified amino acid, a tyrosine, in the core of the protein. To become a fluorophore, though, that tyrosine has to do some really weird (for a protein) chemistry -- it actually self-reacts to form a little pentamer with its own backbone. The amino acids immediately flanking the tyrosine are critical to allowing the reaction to happen, and the rest of the protein stabilizes the reaction even further, to the point that it happens on a reasonable timescale. For a more visual explanation, see http://ift.tt/210HtlD.

Thanks, as so often the case these days, to Anders Knight. And yes, I suspect I will be riding this knowledge gradient until Anders finishes his rotation with us.

February 12, 2016 at 02:53AM

Today I Learned: 1) How to save protocols for later use in Biotek plate reader software. It's the little things.... 2) Lasius neoniger, a common ant in the US and Canada, is commonly known as the labor day ant, because L. neoniger queens fly and mate every year right around labor day. The queen, once mated, finds or digs a promising burrow, then promptly goes into hibernation until the following spring, when it begins laying eggs. 3) The Tesla Model 3 is slated for reveal next month, and is expected to cost between $21,500 and $35,000, depending on what state you live in and what tax incentives you take advantage of.

February 11, 2016 at 02:37AM

Today I Learned: 1) Cyanobacteria (the clade of bacteria best known for their ability to photosynthesize, an early member of which is almost certainly the ancestor of modern chloroplasts) are seriously polyploid, meaning that they have multiple copies of their genome. Different species are reported to have between 3 (Synechococcus elongatus PCC 7942) and hundreds (most other strains) of copies of their chromosome. The numbers also vary quite a bit between exponential phase, when the cells are growing quickly, and stationary phase, when they're nutrient-deprived and more or less at equilibrium. During rapid growth, most species appear to have on the order of 100-300 chromosomal copies, which drops to a few dozen copies during stationary phase. Synechococcus elongatus PCC 7942 (S. elongatus) has been studied rather more carefully. During growth, S. elongatus has... several chromosomal copies (more than three, less than 10). They are spaced evenly throughout the cell and interspaced between carboxysomes, which helps make sure that when they divide, each daughter cell gets roughly the same number of chromosomes. As far as I know, we have no idea why they have so many chromosomal copies. I've been daydreaming for a while about ways to make chromosomes more error-resistant, which would help make synthetic gene circuits more robust. Ultimately, I think you have to do something like having a triplicate genome where every position is checked between the three copies and corrected to the most common version. This seems nigh impossible for many reasons, but one of those reasons was, until today, that bacteria very specifically have exactly one chromosome. Looks like I was wrong on that. Sources: http://ift.tt/23YmxOD http://ift.tt/1XkvVaU 2) ...the difference between the European Union and the Eurozone. The European Union is a... federation...? of countries (in Europe, I hope it goes without saying) which agree to a whole slew of things like open borders with other EU members states, no economic sanctions or tariffs against other EU states, and a bunch of humanitarian standards. The EU also has an elected parliament which is widely considered feckless, as well as a bureaucracy that actually wields some considerable clout internationally. The Eurozone, in contrast, is the subset of the EU which has agreed to use the euro as their standard currency and jointly manages the euro. I know, this is pretty basic stuff, but I've gotta start somewhere. 3) You can make hard cheese from goat's milk. Apparently there is one such cheese (whose name I don't know) made in Greece in very small quantities which is soaked in wine, wrapped in a mix of mud and leaves, and buried in a sandy beach to soak up sea salt for a while before it is sold. Perhaps I should check out how well the goats are treated....

February 10, 2016 at 02:32AM

Today I Learned: 1) There's a step of awesome above automatic watches -- the Atmos atmospheric-pressure-drive clock. This exquisitely constructed clock brand sits in a sealed chamber with ethyl chloride gas. When tuned properly, differences in ambient temperature cause the ethyl chloride pressure to increase and decrease, which winds a spring. It can run "for years" without human intervention. Related to the Atmos is the Beverly Clock, which is a single clock in New Zealand which has operated near-continuously (but not quite continuously) under similar principles without winding since its construction in 1864. While reading about the Atmos, I also stumbled on the idea of the torsional pendulum -- this is a pendulum in which the negative feedback that keeps the pendulum running is torsion rather than gravity. Imagine a pendulum made from string, with a long bar as a head. Now spin the bar. It will wind up the string, then unwind, then anti-wind, then un-anti-wind, and on and on for quite a while. Apparently these are more efficient than traditional pendulums for long-term use, though I don't know why. This TIL fact *also* brought to you by Anders Knight. How long can he keep it up? 2) The player ranking system in Go consists of many levels of kyu (or student ranks), on top of which are nine levels of dan. Today I learned the approximate meaning of a 1-dan difference in skill. Roughly speaking, a player one dan below his opponent should be able to win around 50% of the time when he or she gets an extra free turn at the beginning of the game. That's a big difference. AlphaGo, the Google's headline-making Go-playing program, has beaten the European Go champion, who is a 2-dan professional. It is scheduled to play a 9-dan professional sometime next month. Thanks to Robert Johnson and Chigozie Nri for teaching me this! 3) Apparently doctors don't like electronic record-keeping systems. Really, really don't like them. As in, electronic record-keeping systems are one of the most consistently-complained-about parts of general practitioners' daily work. I find this somewhat flabbergasting.

February 09, 2016 at 01:19AM

Today I Learned: 1) Pseudomonas aeroginosa, a common bacteria known for easily developing antibiotic resistance, can grow in diesel and jet fuel! What an awful way to lose a tank of gas. 2) Automatic watches. They're spring-powered watches that wind themselves from the motion of their bearer. I'll just leave you with these: http://ift.tt/20HcaiZ http://ift.tt/1Xf41Nq http://ift.tt/20HccaR http://ift.tt/1Xf41Ns Anders Knight is on a roll. 3) BglBricks (pronounced "bagel bricks"). If you know what the biobrick standard is, this paragraph is for you (otherwise, see this asterisk*). The BglBrick standard is a modification of the BioBrick standard that lets you put two protein-coding sequences together, and the scar left behind by the restricion site happens to be the codons for a serine-glycine linker, which is the standard flexible linker you put between two domains in a protein fusion. Ok, it's not that it "just happens" to be a serine-glycine linker, it was designed that way. Anyway, it's a cool part standard for making fusion proteins. Don't have a clue how well it works, but it's from Arkin and Dueber, so it can't be all bad. * The biobrick standard is a DNA parts standard. What is a DNA part, you might ask? A DNA part is a string of DNA that encodes something specific and (presumably) useful, like a temperature-sensitive promoter, or a sequence for a fluorescent green protein, or a binding site for a repressor. A part *standard* is usually a specification for what DNA flanks the part, almost always with specific restriction sites to make it easy to clone. The BioBrick standard, developed by Tom Knight for the iGEM synthetic biology competition, is one such standard -- it specifies two restriction sites on each side of the part, in a specific order. The cool thing about the BioBrick standard is that when you cut the pieces with the right parts and ligate them together, you get a *new BioBrick* part that's a composite of the original two, with the right restriction enzymes on either side and no restriction sites in the middle. Unfortunately, the standard BioBrick assembly protocol doesn't actually work very well (it's not clear to me why), so most iGEM teams use other cloning methods. You are now prepared to read the previous paragraph.

February 08, 2016 at 04:39AM

Today I Learned: 1) For all of the advantages of doing work in a Mathematica notebook, they can come out *really* ugly when converted to PDF. 2) Apparently power plants that use water as a liquid medium *actually boil and condense that water*. At least, so says Wikipedia. I find this hard to believe. A little bit of explanation may be in order here. Virtually all power plants (with the exception of some kinds of solar) work by the same basic mechanism -- they spin a really big magnet inside a coil of metal wire, which generates electricity. Moreover, almost all of the power plants that we use* work by heating a liquid or gas, which rises through a turbine, causing it to spin. Usually water is what's used as the liquid/gas. A coal plant, for example, consists of a giant coal-fed fire that heats a giant vat of water, causing it to boil, which makes a whole bunch of steam that drives a turbine. That seems ridiculous to me. Boiling water is one of the easiest ways to lose a ton of energy without really doing much. Why not use something that doesn't take as much energy to boil, like ethanol? Actually, ethanol is a terrible idea, it would probably just catch on fire and/or explode, but surely there's some relatively cheap liquid that is easier to boil and doesn't ignite at power plant temperatures. Surely? Thanks Anders Knight for teaching me this *utterly incomprehensible* "fact". *Exceptions: wind plants spin their turbines directly from the wind; hydroelectric and tidal, which spin their turbines using flowing water. Off the top of my head I can't think of other exceptions. 3) Brown mustard works surprisingly well in lo-mein-like noodles. I was hoping for a cheap immitation of chinese mustard, but I got a really tasty, unexpectedly spicy noodle topping. I may have to try mustard with other noodles now.

February 07, 2016 at 05:06AM

Today I Learned: 1) ...this quick algorithm for cooking dumplings: bring water to a boil; add dumplings; for x = 1:3 let water come to a boil; add a cup of cold water to cool; end for let water come to a boil; remove from water and serve; 2) When making spring rolls, don't despair if they don't get wrapped well, or if they crack a bit, or if they come loose. Fry them carefully and they'll be fine. Maybe. 3) ...how to make a binary counter out of genetic repressors! According to the standard simplified repressor model, at least.

February 06, 2016 at 04:12AM

Today I Learned: 1) ...how to use a Buck converter! That's a circuit component designed to take an input voltage and convert it down to a smaller voltage. Very handy for, oh, say, finding the voltage potential required to evolve oxygen off a catalyst.... This was particularly difficult because I and the high school students involved with this project didn't have any experience with or instructions for a Buck converter, so we had to pretty much guess our way through it. 2) When E. coli get stressed, particularly from starvation, they sometimes transition to an absolutely ridiculous morphology where they keep growing but more or less stop dividing, so they get absurdly long. If you're not familiar with what E. coli normally look like, go do a quick Google image search. Then check out Figure 1 from this paper, in which the authors induced filamentous growth with platinum salts: http://ift.tt/1PqC0xe. Also check out figure 3, where they show what hapens to the filaments when they're moved to platinumless media -- the whole filament shards itself as it apparently undergoes all of its divisions at once. 3) Jonathan Blow, the indie game designer responsible for Braid and, more recently, The Witness, has a fantastic website. It is ancient and ugly, but it has some of the highest density of interesting content I've ever seen on the web. Some of it I've seen before ("The Unreasonable Effectiveness of Mathematics in the Natural Sciences"), most of it I haven't ("The Unreasonable Effectiveness of Mathematics"*). I have not agreed 100% with anything on the site yet, but just about everything has given me a new way of looking at something I'd thought a lot about before. Speaking of which, it's time for me to go into unpaid advertisement mode -- The Witness is really good. Really, really good. I'm not done yet, but if it's going where I think it's going, it's one of the best-executed pieces of psychological manipulation I've seen in the medium. *Not a typo.

February 05, 2016 at 02:26AM

Today I Learned: 1) Anders Knight is on a roll recently -- tonight, he is responsible for information on the Wendelstein (pronounced "vendelshtein"), the Max Plank Institute for Plasma Physics' brand new fusion power research facility. You can get a lot of information here (http://ift.tt/1VTi8GW), but I'll relay some of it anyway. The Wendelstein is a fusion reactor of the stellarator type (for a comparison of stellarators to the more common tokamak designs, see http://ift.tt/1SwHUD7. This is also a really good fast introduction to how a fusion reactor works, in its broadest strokes). It will not produce power. It is instead designed as a research facility to test different configurations of magnetic containment* so we can figure out how to optimally build fusion reactors that might, in a few decades, produce power. One of the cool things about the Wendelstein is how long it can operate -- the facility has already performed runs as long as a quarter second, and plan to ramp up to thirty seconds. For a fusion reactor, this is an eternity. Another cool thing about the Wendelstein is the insane exactness required to build it. For example, each of the 50 non-planar (I'm not sure what that means) superconducting coils of the Wendelstein's magnetic containment system measures roughly 3.5m x 2.5m x 1.5m and is machined to a precision of *three millimeters*. Speaking of the magnetic containment system, another cool thing about the Wendelstein is how *awesome* it looks. I mean, check this out: http://ift.tt/1VTi8GY. That's not science fiction, that's engineering that's being done *right now*. * Fusion reactions only happen at super-high temperatures, temperatures which are hot enough to destroy literally any known physical material you put them in. Fusion reactors contain their plasma with complex magnetic traps. You know in Star Trek when ships lose magnetic containment and explode? That's because the plasma in their fusion reactors escapes the reactor and briefly exposes the rest of the ship to temperatures not dissimilar to the interior of the sun. 2) When E. coli cultures are infected with bacteriophages (viruses that attack bacteria), it apparently often causes the culture to form visible clumps or filaments. There's also some rumor that phage-infected cultures tend to get really foamy, from all the protein released by dying cells. 3) In a drug (or medical device) trial, each subject costs, on average, about $10,000. The total cost of passing regulation for a medical device averages around $150 million. Source: a lecture by a senior guy from Medtronic.

February 04, 2016 at 02:31AM

Today I Learned: 1) Big news! http://ift.tt/1VOYCLF 2) Something to speed up PCR linearizations, or really any amplification you have to check on a gel -- start DNA purification while the gel runs, before you know what needs to be purified and what doesn't. It'll cost more, because you end up prepping some samples that didn't turn out anyway, but it saves some time. I'm pretty sure I knew this at one point (iGEM, can you confirm?), but I learned it anew today. 3) Hexane is a surprisingly cheap solvent. It also comes in a LOT of purities: http://ift.tt/1JZCX2F

February 03, 2016 at 02:12AM

Today I Learned: 1) Let me tell you the story of FDC Willard, one of the most famous published authors in low-temperature physics. The story of FDC Willard is inextricably bound to the story of Jack Hetherington, a low-temperature physicsist. In 1975, he submitted a paper of his for review, which happened to be a single-author paper. He learned from a friend of his during reviews that his paper would be rejected on editorial grounds because he had used the standard "we" throughout the paper, which was in this case improper as there was only one author. Jack didn't want to completely re-work the paper, so he decided to add a second author instead. Apparently having no colleagues in low-temperature physics who he wanted to add to his paper, he decided instead to add his cat, Felix domesticus cattus Willard, i.e. FDC Willard. The paper was accepted. Willard would later go on to first-author a paper on... well, frankly I couldn't decipher what it was about, but it was another low-temperature physics paper having something to do with helium 3. Thanks to Anders Knight on this one! 2) Enzymes are really useful for a lot of industrial chemistry. It's relatively common to mix up a big reaction vessel with some substrate, pour in a bunch of enzyme, and stir it around or heat it to produce a reaction, and the enzyme greatly reduces the energy required to do so. The product can be extracted, sold, and thus money is made. However, enzymes are expensive to produce. You really don't want to make a giant reaction vessel's worth of enzyme and then throw it away. You can filter out enzymes with some sort of antibodies or similar trick, but that tends to be expensive and inefficient. A generally better way to re-use enzymes (as I learned today) is to not just dump the enzymes in -- instead, covalently bond them to some solid thing, like a rod or paddle or mixing aparatus, which you can stick in the reaction vessel... and just pull out when you're done, ready to be re-used. Enzymes used this way are referred to as immobilized biocatalists. The tricky part is figuring how to attach the enzyme without destroying its function. Rumor has it that one surprisingly effective way to attach enzymes to a surface is... to bake them on! That's it. Just heat them a bunch. Apparently works better than it should, although I don't have a reference for that. Credit also goes to Anders Knights on this one. 3) Otto E. Rössler, inventor of the Rössler atractor (a rather famous system of simple differential equations with rather complex properties that's studied as a cannonical example of chaotic dynamics -- check out the wiki page for some cool snapshots of the system) is one of the most wide-ranging scientists I've ever encountered, and possibly one of the craziest. For example, he is one of the scientists most against the experiments in the Large Hadron Collider on the grounds that it could produce Earth-destroying black holes. Check out his blog, linked below. Among other things, it covers: cryodynamics (a "sister theory to thermodynamics"); ebola; the LHC; Interstellar; the ethics of curing elephant autism. http://ift.tt/1mc997p

February 02, 2016 at 03:53AM

Today I Learned: 1) Our best images of single stars (other than our own) come from IR interferometer telescopes. IR means it collects infra-red light, presumably because it... actually, I don't know why IR would be better than visible light. In any case, an interferometer is a telescope array that collects the same image and compares them from slightly different angles*. Anyway, one of the IR interferometers has captured some pictures of other stars that are, like, more than one pixel! We're talking about images on the order of a couple dozen pixels square. It turns out that a lot of the big ones (which are easiest to image) are really squashed-looking -- they rotate so fast that they bulge at the center. This effect happens in all spinning bodies, but on the ones in our solar system the effect is small. *Apparently the total resolution you get is the sum of the resolutions of the individual telescopes, and it's easier to build a bunch of small telescopes than it is to build one big one. You can probably also split up an array to do a bunch of low-resolution tasks at once, though this is just speculation on my part. 2) Glassblowing, as a field, has a lot of really archaic verbage. Here are a couple of terms I learned today (I learned them verbally so the spelling is virtually guaranteed to be wrong). The puntee is the rod on which you gather glass. Puntees can be solid, for making solid glass pieces, or hollow, so you can blow into them to create things with holes or big rounded cavities. Gathering glass, by the way, is what you do when you put a blob of glass on the end of a puntee. The blob of glass on the end of the puntee is the moil. One of the common things one does with a moil is to roll it on a mauvre, which is a plate traditionally made of marble (from which the name comes), but now is more likely... steel? I'm fuzzy on this. In any case, you roll a moil around on the mauvre to shape it. Thanks to Tasha Johnson for teaching me these! 3) ...what a Luer lock is. Basically, it's a standard of screw fitting for plastic parts like syringe tips and tubing. I've been using them for a while now but I just now learned what they're called.

February 01, 2016 at 03:01AM

Today I Learned: 1) So thalidamide is this decently effective drug whose enantiomer is highly teratogenic, which led to tens of thousands of severe birth defects when the drug was first administered as a painkiller for pregnant women(!). It's always struck me that this was an awfully sloppy mistake to make. After all, chirality is something chemists know about, and presumably it would have come up during the synthesis of thalidamide that there was an enantiomer they should really deal with. Wy did anyone think it was a good idea to administer both isomers of a drug? Well, it turns out that it wasn't just sloppy chemistry -- thalidamide can *switch its own racemization*. If you make a pure mix of 100% of either thalidamide enantiomer, it will decay back to a 50/50 over time, *especially* in biological conditions. I had no idea there were compounds that could do that. (I'm actually not sure whether the original thalidamide problem was because of this, or because the chemists that made it were sloppy. This bears more research) 2) Here's a bit of a thought experiment I've been thinking about for a while. Take a box full of salt water (say, sodium chloride). Apply an electric field across the water by putting a positively-charged rod at one end and a negatively charged rod at the other. The sodium ions should be attracted to the negative pole, and the chloride ions should be attracted to a negative pole, so this will create a gradient of charged ions across the box. Now spoon out a bunch of the water at one end, near the pole, and seal it away in a jar. Do you now have a bunch of charged water? It seems like you really, really ought to, but the idea of water with a net charge seems really crazy to me... ...well, today I was reminded that we make charged water for a bunch of reasons using a technique called electrospray. There, though, you're putting charge on a bunch of water droplets and spitting them out using an electromagnet. It turns out that if you put TOO much charge on a droplet, the repulsive forces of the charge overcome the cohesive forces of the water and it explodes. There's a special name for this explosion -- a Coulomb explosion. 3) ...not to let the lab turbidostat sit on its own for more than a day, or it starts to spew slightly contaminated media all over the bench. My secondary containment system (a box) worked decently well at limiting the damage, at least....