Probably, but a lot of the contemporary stuff I've read about Goldwater was pretty scary. To give the man some credit, he was an idealist, not necessarily a bad guy - but in retrospect he sounds disturbingly like the modern Green party and other elements of the socialist left, who propose a radical policy without fully considering the potential effects. He also didn't do much (that I'm aware of) to dissociate him from his fruitier supporters; loudly disavowing Southern racism in its entirety (while maintaining that federal legislation was an inappropriate way to address it) would have alienated half the people who voted for him.
but not, it needs to be said, nominated by conservatives.
I sort of agree, but I'm also kind of curious how you define "conservative". Most Americans who label themselves that way revere Reagan as a near-deity. By any modern standard, he's about as close as you get to a true conservative. Of course the term has been abused so much that no one really knows for sure what it means any more. The Republican party has often been described as a cross between social conservatives and fiscal conservatives, but even these labels tend to be very inconsistent. (Remember, the budget was briefly balanced under a "liberal" president, and now a "conservative" has fucked it again.)
You could call the Libertarians "conservative", but only in the anti-government sense. Most libertarians I've known were atheists and definitely not traditionalist.
How fast is this gui OS running on the atari, compared to current desktops that are well over 1000x faster that run Gnome/KDE/WinXP ?
Well, shit, judging from my recent experiences with all three of these on *fast* hardware, it couldn't possibly be worse.
Joking aside, as much of a waste of time these projects seem like it's still impressive how they can do so much with so little. The evolution of the desktop environment seems to have been entirely based on development of processor-intensive eye-candy for the past five or so years, with the result being that each new system makes the latest hardware actually seem slower for interactive tasks than what came before. I've yet to use another computer that feels as fast as my Mac Centris 610 did back in 1993, running System 7.1. I've also heard amazing things about the late-model Amigas. There seems to be virtually no serious work done optimizing desktop software, with the exception of Aqua (and that only because the initial releases were too slow even by today's pathetic standards).
I'm still using WindowMaker on Linux, because I find it has exactly the feature set I need (no desktop! yay!), and because since it hasn't been developed much for years it doesn't get progressively slower.
You're not licensing a television, you're buying a license to have a TV set in your house
Actually, this is probably worse. Sorry, but this is like the ultimate incarnation of the "nanny state", when the government can fine you for illegally watching a public broadcast in the privacy of your home. In America, of course, we all pay for PBS whether we want to or not (I don't own a TV), but I would much prefer to have this extracted from my tax dollars invisibly than having to file a fucking license in order to avoid nasty visits from the BBC cops. These assclowns actually think it's funny.
I hate to sound like another hysterical Slashbot equating Bush with the rise of Naziism, but if we ever end up in a society like 1984, it will be because of unchecked excesses of this sort. "Brazil" doesn't seem quite as absurd once you've read about the TV licensing goons. The Brits I know (some very close friends) understand exactly how absurd this is, too.
To be fair, though, is this necessarily the fault of the company? If you're from Europe, you're well aware that Byzantine government regulations are every bit as much in style as over here in the States, although individual examples vary widely in absurdity. It could be required thanks to some amendment that one of your Green politicians pushed through for God-knows-what reason. (Although this wouldn't really be any less absurd than a plant EULA.) You only have to look at the power-line/EMF scare in the US to understand the ill effects of well-meaning zealotry combined with utter ignorance.
In addition, the science behind Hubble was realistic to begin with. None of this microgravity ant farm bullshit. As someone else pointed out, the Columbia disaster is even more tragic when you consider how worthless many of the mission goals were - seven people died for an elementary school science fair project.
Oh thank you thank you thank you. Someone else understands.:)
Seriously, I'm all for sending probes to every object in the solar system - I think exploration of the Jovian moons (especially Europa) should be the primary goal of our space program. The ISS is being justified as a research tool but its actual uses are minimal, and it's really just a bone thrown to contractors and an excuse to keep the shuttles running.
2) Congress people too stupid to care about space.
The scientific justifications for the ISS were primarily bullshit in the first place. If Congress weren't so useless much of the time, they'd never have poured money into it to begin with; that they did suggests that either they really are stupid, but not the way you meant, or that they voted for it to make the defense contractors happy.
I work in a field which is being touted as one of the reasons why we need space-based research, and it just makes me hyperventilate when I think about how much money has been wasted on attempts to realize this. In the meantime, the ground-based techniques have accelerated immensely to the point at which they're far more sophisticated than anything we could put in the ISS. Very few people in the field take this garbage seriously, but laymen don't know enough to check out the facts and it's good PR copy for NASA to say that space research has medical applications.
Good for you, champ. I'm sure you're one l33t motherfucker where Windows administration is concerned, but you're not going to win this argument on anecdotal evidence. I can quote just as many people (myself included) who've had miserable performance and stability problems with XP on new hardware. This proves nothing. (There are miserable stability and performance problems with Linux too, but I stick with Linux because my experience with Windows makes it seem like Microsoft intentionally designed it to prevent me from getting any work done. At least it will let me customize stuff, rather than inflicting Microsoft's pathetic focus-group-driven version of "innovation" on me.)
As for Windows operability on corporate nets. . . yeah, I'm one of those "Linux groupies". I didn't bother learning Windows because it wasn't in my job description to show some postdoc how to set up RealPlayer so he could watch Fox News. The real problem in my environment, where all software development was done on Linux, was with people who were supposed to be technically proficient but were too lazy to learn to use anything other than Windows, and who suffered from the delusion that it made them more productive. After having to support people whose idea of a development environment was XP + Linux-on-VMWare + Exceed, I've decided that I will never, ever let myself get involved in anything involving Windows support.
As for outdated comparisons, fair enough. However, virtually every argument I heard against Linux at my old job was based on exactly these out-of-date misconceptions, spouted by people who thought XP was the greatest thing ever but whose only experience with a Linux desktop was using a 486 running RedHat 5 with crappy video drivers. Macs pretty much got the same treatment; funny thing was, I almost never heard speed as an argument against Macs - it was always compatability or stability. This from Windows users! Sheesh.
I bought my laptop to do actual science on - I needed something with a big screen and reasonably powerful graphics. In my current lab, it turns out that I'm usually better off using my laptop than the lab computers; it's faster, and I don't have to compete with four other groups for CPU time. So, yeah, I need a lot of memory too; I can tear through 512 pretty quickly. My other computer is a relatively ancient SGI workstation, okay for web browsing/graphics/light coding, but not much for number crunching.
The problem with a laptop, however, is that bottlenecks seem to crop up very quickly. At my old job, I had a dual Xeon with more than a gig of RAM, several Ultra 160 disks on multiple channels, and a Quadro card, and there was virtually nothing that could slow it down; I could run huge simulations or render images on one processor and still use the other for web browsing or lighter work. On my laptop, any heavy-duty task will render the entire system inoperable.
Do you remember school? Remember being slammed into lockers and called "geek" or "fag" or "Poindexter" by the jocks? Remember being told to play dumb, or to dress differently, or to "get a life"? I do.
Point being, this sort of bullshit is exactly why the Pledge needs to be eliminated altogether, in my opinion. Nothing that goes on in an official context is ever truly "voluntary". I'm all for students or teachers meeting privately, outside of instruction time, for Bible study or whatever- as far as I'm concerned, they're just another club. They can use school grounds, and I don't see a problem with them advertising as long as they're not given preferential treatment. BUT, once this enters into class, it immediately becomes coercion. The student who chooses to exercise his or her constitutional right is the one getting a wedgie on the bus ride home.
Anyway, this is most definitely "News for Nerds", because we tend to be very consciously non-conformist and thus the target of these ridiculous indoctrination efforts. I never minded the religious overtones when I was in the Boy Scouts or at (private) college, because I was there by choice and others were not paying for me. To feel threatened for speaking out in publically-funded, mandatory primary school, however, is offensive.
(I also oppose the pledge because I think it's a spectacularly crappy idea; we should be above loyalty oaths. I'd rather see it exchanged for mandatory national service of some sort - not just military, of course - with almost no exemptions.)
Some funding groups, like HHMI, will also defray costs. Regardless, as has been mentioned above, paying to have your article published (especially if you want color figures) is not all that uncommon with conventional journals too.
Computer scientists, being handy with the web, typically publish their papers and then put them up on their websites, playing "civil disobedience."
Hmmm, I've seen plenty of biologists who do this, including my former boss, and none of them have been busted for this (so far). However, these usually tend to be fairly computer-savvy people already.
Anyway, PubMed is already a much better way to find publications of interest in biomedical research than any other mechanism, since it's very well curated and comprehensive. The problem, of course, is that once you find an article you still can't get the full text unless you have an online subscription. I've had to go through my old employer to download Cell papers because my school doesn't want to shell out several million dollars to Elsevier. (Which is one of the reasons why one of the profs in my department is one of the PLoS founders.)
By the way, biological journals, at least the ones I've dealt with, give you a fat stack of glossy reprints free of charge - anything more than that, of course, will cost a shitload. I have a folder with about four copies of each of my articles, but I'd like to have the full issues too.
1. And, yes, I am against unrestrained free trade. It doesn't work. Example: We put laws in place to protect the environment and U.S. firms move manufacturing jobs to other countries where pollution laws are lax or non-existent.
I agree with this part. This is why organizations such as the WTO are important, because they can (in theory) help to insure uniform standards.
2. You capitalist bastards think companies should only be concerned with maximizing profits but that workers who are only concerned about maximizing their wages are somehow evil. Explain that logic to me.
Well, I tend to agree with the person you're replying to, and this doesn't describe my beliefs (or his, I suspect) at all. I do not think companies should only be concerned with maximizing profits; however, I do not feel that companies owe me or society anything (beyond a certain minimum of reasonable behavior, which goes back to point 1). What I object to is employees (or companies; it goes both ways in the US, and both are deplorable) who expect to keep their jobs under any circumstances, and expect the government to keep them employed. Like the steelworkers who pushed for a disasterous tariff rather than face the facts and modernize their industry.
By the way, if you're really a 42-year-old engineer with 2 decades of experience, your income is probably in the top few percent of American households, so stop whining about capitalism. I'd prefer to have my hard work and skill determine my salary, not the government.
4. Just what is the typical software engineer supposed to do that will make up for a 20/1 cost ratio between himself and his Indian counterparts?
[etc.]
In all cases, nothing. Tough shit. Find another job, retrain if necessary. Just because you've trained for a job in the technology industry does not mean that the technology industry automatically owes you a job. Alternately, settle for a lower standard of living, and accept a job with a salary more in line with what the rest of the world makes. Better yet, stop whining to the government, and organize a boycott of companies that outsource extensively. I'll join that, very enthusiastically.
Folding@Home is a great cause. We all know this. Finding a cure for a terrible disease is very noble.
I'm very curious about how and why you think these two are related. If you'd actually read the papers that have resulted from this project, you'd find that they're doing really interesting theoretical work that bears little relation to human disease. It's worthy research, but the link to curing Alzheimer's or BSE is tenuous at best - it just makes it seem more relevant.
I do agree it's a more useful application of spare cycles than Seti@Home, although this isn't exactly difficult. Let's just not delude ourselves about the power of computational biology.
1) Stanford is not public, though it receives large amounts of public money for various research activities (like its particle accelerators or its entire biomedical sciences program).
2) Yes, public universities do sometimes end up in corporate partnerships. I've heard of several cases- I think one of the UC schools has a deal with a large pharma company.
If it makes you feel any better, the results that protein-folding simulations provide aren't very useful for genetic engineering or commercial applications. They're simply interesting tests of theory, but they have nearly zero application to human medical sciences. Unfortunately the press coverage of computational protein structure simulation (thank you, Slashdot) tends to make overbroad generalizations about how we'll replace experimentalists with supercomputers and do all our biology in silico. In a word, horseshit.
There is definitely a place for this type of work, particularly when coupled with biophysical studies. But the idea that we'll be able to determine the structure of a protein from its sequence is laughable.
If Rockefeller admissions work the way I think they do, interviewing means that you had the opportunity to attend. Rockefeller is a posh place with excellent science. Are you working with ion channels now?
I was already admitted when I visited. However, I didn't go there, for a variety of reasons; among others, there was no guarantee that I'd get to work in that lab, so it would have been foolish to go there solely because of one professor. I was mainly interested (and still am) in doing structural biology - I just found ion channels particularly interesting. (Although, personally, I like the membrane pumps better.)
As for the hypotheses: one of the principal paradoxes of the potassium channel was how it maintains selectivity, and blocks sodium which actually has a smaller radius. The pore is negatively charged and sized exactly right so that a dehydrated potassium ion is stabilized, but a sodium is too small for the same effect. Thus sodium remains hydrated on the other side of the membrane, while potassium zips through. This was guessed at before, but the structure illustrated it exactly. He won the Lasker award for that alone, but the other structures- several of which detail gating mechanisms- have been equally cool.
Having fitted a number of experimental 3A maps, I can unequivocally say that the global structure is far from obvious
Okay, "obvious" was overstated. I guess I meant it relative to protein-folding simulations, which are beyond useless for telling you the tertiary structure. 3A is still enough to indicate the overall fold, i.e. "global structure", even if sidechains are incomprehensible.
CNS/CNX (the current versions of the molecular dynamics refinement program) works very well on multiprocessor systems, using OpenMP.
That's news to me - I've looked at the CNS source code and I couldn't find any parallel routines. I've seen stuff about using compiler auto-parallization, but this doesn't seem very effective. CNX may be parallel, but you have to shell out thousands of dollars to Accelrys for the priviledge of using it, and most academic labs would prefer to use the free version.
Regardless, I suspect the problems introduced by distributed computing would make it completely unfeasible for this type of work.
I am wondering whether he used distributed or parallel protien folding simulations for some of his work...
As the AC pointed out, none whatsoever. Almost nobody doing crystallography does, in fact. It continues to amaze me how many people think that simulations are going to replace experiment.
X-ray crystallography on globular protiens usually yields poor results (it is hard to get the X-rays to diffract to show the inner channel structure) compared to crystalline/regular protiens.
This is incorrect: crystallography of membrane proteins is extremely difficult compared to globular proteins. However, the techniques are getting better and better and there are now a number of very diverse membrane protein structures available.
Pumps are different. There is also a P-type ATPase, the calcium pump from muscle tissue, and the F1F0 ATPase, which pumps protons in the mitochondria. John Walker won the Nobel for the latter structures (although most of these were non-membrane). One of my former co-workers solved the cytochrome BC1 oxidase, also very large.
I think (can't remember for sure) that one reason the channels are so difficult is that they're nowhere near as stable when you take them out of the membrane. Intuitively, this makes sense comparing the structures: most of the others have coherent helical bundles, but the potassium channel does not. The key is allowing a very high rate of transport while maintaining excellent specificity.
I am willing to bet that it or other distributed computing projects are actually quite critical in the types of work represented by this Nobel prize.
Nope. Refinement of structures often uses molecular dynamics, one of the classical simulation methods and also a (very slow) way of looking at protein folding. However, the software that does this is single-processor and actually doesn't require too much more power than a fast desktop. These structures were all at around 3-Angstrom resolution, and once you're able to collect and properly phase X-ray data of that quality, the global structure is obvious.
Everything that was known or guessed about membrane channel structures before this work was done using electrophysiology and classical molecular biology - purely experimental methods, both quantitative and qualitative. Relatively few structural biologists use standalone simulations, and this is only ever done after a high-resolution structure has already been obtained. The only people using de novo simulations are those working on protein folding (most of whom are actually experimentalists) or protein design people.
I know of no important native biological structure solved with the aid of prior simulation; this is not to say that none exist, but you're grossly overestimating the importance of theoretical methods.
There was pretty much no doubt that MacKinnon would win it eventually - but it's a bit surprising that it came so soon, considering he's at the height of his career. He's only published four papers this year, but they're all Science or Nature (including one cover article). We can probably expect equally terrific work from him in the future.
I interviewed with him earlier this year (I applied to Rockefeller largely because of his lab), and he's one of the most intensely brilliant people I've ever met. There are very few scientists who will master a completely different technique in the middle of their career, while working on the same area of research. Fewer still are able to dominate the field. When I took physiology in college, we read multiple articles which described hypotheses proved by a single figure in one of MacKinnon's papers.
(There are actually an increasing number of membrane protein structures available, some of them quite large. However, ion channels are apparently especially difficult to study, and none were solved before MacKinnon started.)
It actually varies more than you'd expect. Kary Mullis didn't have to wait too long after PCR was invented, because the technique was immediately useful and revolutionary. Michel/Huber/Diesenhofer won in 1988 for X-ray structure of the photosynthetic reaction center, which they'd published only a few years previously. Until then, people had thought high-res structures of membrane proteins couldn't be determined.
Scientific research is weird, because what may in retrospect appear to be a "revolutionary" concept may take years to be recognized as such, even if they're immediately seen to be somewhat useful. And of course there are cases where a discovery was initially dismissed entirely, and the brilliance of those responsible for it wasn't realized until long after. Catalytic RNA and prions are the best recent example I can think of.
One of the places she worked at, Birkbeck College in London, suggests that she missed out on multiple prizes - after DNA, she worked on viral strucutres with Aaron Klug, who eventually won the Nobel himself.
Slashdot Mirror
AuH2O was the last true conservative in the GOP.
Probably, but a lot of the contemporary stuff I've read about Goldwater was pretty scary. To give the man some credit, he was an idealist, not necessarily a bad guy - but in retrospect he sounds disturbingly like the modern Green party and other elements of the socialist left, who propose a radical policy without fully considering the potential effects. He also didn't do much (that I'm aware of) to dissociate him from his fruitier supporters; loudly disavowing Southern racism in its entirety (while maintaining that federal legislation was an inappropriate way to address it) would have alienated half the people who voted for him.
but not, it needs to be said, nominated by conservatives.
I sort of agree, but I'm also kind of curious how you define "conservative". Most Americans who label themselves that way revere Reagan as a near-deity. By any modern standard, he's about as close as you get to a true conservative. Of course the term has been abused so much that no one really knows for sure what it means any more. The Republican party has often been described as a cross between social conservatives and fiscal conservatives, but even these labels tend to be very inconsistent. (Remember, the budget was briefly balanced under a "liberal" president, and now a "conservative" has fucked it again.)
You could call the Libertarians "conservative", but only in the anti-government sense. Most libertarians I've known were atheists and definitely not traditionalist.
How fast is this gui OS running on the atari, compared to current desktops that are well over 1000x faster that run Gnome/KDE/WinXP ?
Well, shit, judging from my recent experiences with all three of these on *fast* hardware, it couldn't possibly be worse.
Joking aside, as much of a waste of time these projects seem like it's still impressive how they can do so much with so little. The evolution of the desktop environment seems to have been entirely based on development of processor-intensive eye-candy for the past five or so years, with the result being that each new system makes the latest hardware actually seem slower for interactive tasks than what came before. I've yet to use another computer that feels as fast as my Mac Centris 610 did back in 1993, running System 7.1. I've also heard amazing things about the late-model Amigas. There seems to be virtually no serious work done optimizing desktop software, with the exception of Aqua (and that only because the initial releases were too slow even by today's pathetic standards).
I'm still using WindowMaker on Linux, because I find it has exactly the feature set I need (no desktop! yay!), and because since it hasn't been developed much for years it doesn't get progressively slower.
You're not licensing a television, you're buying a license to have a TV set in your house
Actually, this is probably worse. Sorry, but this is like the ultimate incarnation of the "nanny state", when the government can fine you for illegally watching a public broadcast in the privacy of your home. In America, of course, we all pay for PBS whether we want to or not (I don't own a TV), but I would much prefer to have this extracted from my tax dollars invisibly than having to file a fucking license in order to avoid nasty visits from the BBC cops. These assclowns actually think it's funny.
I hate to sound like another hysterical Slashbot equating Bush with the rise of Naziism, but if we ever end up in a society like 1984, it will be because of unchecked excesses of this sort. "Brazil" doesn't seem quite as absurd once you've read about the TV licensing goons. The Brits I know (some very close friends) understand exactly how absurd this is, too.
To be fair, though, is this necessarily the fault of the company? If you're from Europe, you're well aware that Byzantine government regulations are every bit as much in style as over here in the States, although individual examples vary widely in absurdity. It could be required thanks to some amendment that one of your Green politicians pushed through for God-knows-what reason. (Although this wouldn't really be any less absurd than a plant EULA.) You only have to look at the power-line/EMF scare in the US to understand the ill effects of well-meaning zealotry combined with utter ignorance.
In addition, the science behind Hubble was realistic to begin with. None of this microgravity ant farm bullshit. As someone else pointed out, the Columbia disaster is even more tragic when you consider how worthless many of the mission goals were - seven people died for an elementary school science fair project.
Oh thank you thank you thank you. Someone else understands. :)
Seriously, I'm all for sending probes to every object in the solar system - I think exploration of the Jovian moons (especially Europa) should be the primary goal of our space program. The ISS is being justified as a research tool but its actual uses are minimal, and it's really just a bone thrown to contractors and an excuse to keep the shuttles running.
2) Congress people too stupid to care about space.
The scientific justifications for the ISS were primarily bullshit in the first place. If Congress weren't so useless much of the time, they'd never have poured money into it to begin with; that they did suggests that either they really are stupid, but not the way you meant, or that they voted for it to make the defense contractors happy.
I work in a field which is being touted as one of the reasons why we need space-based research, and it just makes me hyperventilate when I think about how much money has been wasted on attempts to realize this. In the meantime, the ground-based techniques have accelerated immensely to the point at which they're far more sophisticated than anything we could put in the ISS. Very few people in the field take this garbage seriously, but laymen don't know enough to check out the facts and it's good PR copy for NASA to say that space research has medical applications.
Good for you, champ. I'm sure you're one l33t motherfucker where Windows administration is concerned, but you're not going to win this argument on anecdotal evidence. I can quote just as many people (myself included) who've had miserable performance and stability problems with XP on new hardware. This proves nothing. (There are miserable stability and performance problems with Linux too, but I stick with Linux because my experience with Windows makes it seem like Microsoft intentionally designed it to prevent me from getting any work done. At least it will let me customize stuff, rather than inflicting Microsoft's pathetic focus-group-driven version of "innovation" on me.)
As for Windows operability on corporate nets. . . yeah, I'm one of those "Linux groupies". I didn't bother learning Windows because it wasn't in my job description to show some postdoc how to set up RealPlayer so he could watch Fox News. The real problem in my environment, where all software development was done on Linux, was with people who were supposed to be technically proficient but were too lazy to learn to use anything other than Windows, and who suffered from the delusion that it made them more productive. After having to support people whose idea of a development environment was XP + Linux-on-VMWare + Exceed, I've decided that I will never, ever let myself get involved in anything involving Windows support.
As for outdated comparisons, fair enough. However, virtually every argument I heard against Linux at my old job was based on exactly these out-of-date misconceptions, spouted by people who thought XP was the greatest thing ever but whose only experience with a Linux desktop was using a 486 running RedHat 5 with crappy video drivers. Macs pretty much got the same treatment; funny thing was, I almost never heard speed as an argument against Macs - it was always compatability or stability. This from Windows users! Sheesh.
I bought my laptop to do actual science on - I needed something with a big screen and reasonably powerful graphics. In my current lab, it turns out that I'm usually better off using my laptop than the lab computers; it's faster, and I don't have to compete with four other groups for CPU time. So, yeah, I need a lot of memory too; I can tear through 512 pretty quickly. My other computer is a relatively ancient SGI workstation, okay for web browsing/graphics/light coding, but not much for number crunching.
The problem with a laptop, however, is that bottlenecks seem to crop up very quickly. At my old job, I had a dual Xeon with more than a gig of RAM, several Ultra 160 disks on multiple channels, and a Quadro card, and there was virtually nothing that could slow it down; I could run huge simulations or render images on one processor and still use the other for web browsing or lighter work. On my laptop, any heavy-duty task will render the entire system inoperable.
Do you remember school? Remember being slammed into lockers and called "geek" or "fag" or "Poindexter" by the jocks? Remember being told to play dumb, or to dress differently, or to "get a life"? I do.
Point being, this sort of bullshit is exactly why the Pledge needs to be eliminated altogether, in my opinion. Nothing that goes on in an official context is ever truly "voluntary". I'm all for students or teachers meeting privately, outside of instruction time, for Bible study or whatever- as far as I'm concerned, they're just another club. They can use school grounds, and I don't see a problem with them advertising as long as they're not given preferential treatment. BUT, once this enters into class, it immediately becomes coercion. The student who chooses to exercise his or her constitutional right is the one getting a wedgie on the bus ride home.
Anyway, this is most definitely "News for Nerds", because we tend to be very consciously non-conformist and thus the target of these ridiculous indoctrination efforts. I never minded the religious overtones when I was in the Boy Scouts or at (private) college, because I was there by choice and others were not paying for me. To feel threatened for speaking out in publically-funded, mandatory primary school, however, is offensive.
(I also oppose the pledge because I think it's a spectacularly crappy idea; we should be above loyalty oaths. I'd rather see it exchanged for mandatory national service of some sort - not just military, of course - with almost no exemptions.)
Some funding groups, like HHMI, will also defray costs. Regardless, as has been mentioned above, paying to have your article published (especially if you want color figures) is not all that uncommon with conventional journals too.
Computer scientists, being handy with the web, typically publish their papers and then put them up on their websites, playing "civil disobedience."
Hmmm, I've seen plenty of biologists who do this, including my former boss, and none of them have been busted for this (so far). However, these usually tend to be fairly computer-savvy people already.
Anyway, PubMed is already a much better way to find publications of interest in biomedical research than any other mechanism, since it's very well curated and comprehensive. The problem, of course, is that once you find an article you still can't get the full text unless you have an online subscription. I've had to go through my old employer to download Cell papers because my school doesn't want to shell out several million dollars to Elsevier. (Which is one of the reasons why one of the profs in my department is one of the PLoS founders.)
By the way, biological journals, at least the ones I've dealt with, give you a fat stack of glossy reprints free of charge - anything more than that, of course, will cost a shitload. I have a folder with about four copies of each of my articles, but I'd like to have the full issues too.
1. And, yes, I am against unrestrained free trade. It doesn't work. Example: We put laws in place to protect the environment and U.S. firms move manufacturing jobs to other countries where pollution laws are lax or non-existent.
I agree with this part. This is why organizations such as the WTO are important, because they can (in theory) help to insure uniform standards.
2. You capitalist bastards think companies should only be concerned with maximizing profits but that workers who are only concerned about maximizing their wages are somehow evil. Explain that logic to me.
Well, I tend to agree with the person you're replying to, and this doesn't describe my beliefs (or his, I suspect) at all. I do not think companies should only be concerned with maximizing profits; however, I do not feel that companies owe me or society anything (beyond a certain minimum of reasonable behavior, which goes back to point 1). What I object to is employees (or companies; it goes both ways in the US, and both are deplorable) who expect to keep their jobs under any circumstances, and expect the government to keep them employed. Like the steelworkers who pushed for a disasterous tariff rather than face the facts and modernize their industry.
By the way, if you're really a 42-year-old engineer with 2 decades of experience, your income is probably in the top few percent of American households, so stop whining about capitalism. I'd prefer to have my hard work and skill determine my salary, not the government.
4. Just what is the typical software engineer supposed to do that will make up for a 20/1 cost ratio between himself and his Indian counterparts?
[etc.]
In all cases, nothing. Tough shit. Find another job, retrain if necessary. Just because you've trained for a job in the technology industry does not mean that the technology industry automatically owes you a job. Alternately, settle for a lower standard of living, and accept a job with a salary more in line with what the rest of the world makes. Better yet, stop whining to the government, and organize a boycott of companies that outsource extensively. I'll join that, very enthusiastically.
Folding@Home is a great cause. We all know this. Finding a cure for a terrible disease is very noble.
I'm very curious about how and why you think these two are related. If you'd actually read the papers that have resulted from this project, you'd find that they're doing really interesting theoretical work that bears little relation to human disease. It's worthy research, but the link to curing Alzheimer's or BSE is tenuous at best - it just makes it seem more relevant.
I do agree it's a more useful application of spare cycles than Seti@Home, although this isn't exactly difficult. Let's just not delude ourselves about the power of computational biology.
Two points:
1) Stanford is not public, though it receives large amounts of public money for various research activities (like its particle accelerators or its entire biomedical sciences program).
2) Yes, public universities do sometimes end up in corporate partnerships. I've heard of several cases- I think one of the UC schools has a deal with a large pharma company.
If it makes you feel any better, the results that protein-folding simulations provide aren't very useful for genetic engineering or commercial applications. They're simply interesting tests of theory, but they have nearly zero application to human medical sciences. Unfortunately the press coverage of computational protein structure simulation (thank you, Slashdot) tends to make overbroad generalizations about how we'll replace experimentalists with supercomputers and do all our biology in silico. In a word, horseshit.
There is definitely a place for this type of work, particularly when coupled with biophysical studies. But the idea that we'll be able to determine the structure of a protein from its sequence is laughable.
If Rockefeller admissions work the way I think they do, interviewing means that you had the opportunity to attend. Rockefeller is a posh place with excellent science. Are you working with ion channels now?
I was already admitted when I visited. However, I didn't go there, for a variety of reasons; among others, there was no guarantee that I'd get to work in that lab, so it would have been foolish to go there solely because of one professor. I was mainly interested (and still am) in doing structural biology - I just found ion channels particularly interesting. (Although, personally, I like the membrane pumps better.)
As for the hypotheses: one of the principal paradoxes of the potassium channel was how it maintains selectivity, and blocks sodium which actually has a smaller radius. The pore is negatively charged and sized exactly right so that a dehydrated potassium ion is stabilized, but a sodium is too small for the same effect. Thus sodium remains hydrated on the other side of the membrane, while potassium zips through. This was guessed at before, but the structure illustrated it exactly. He won the Lasker award for that alone, but the other structures- several of which detail gating mechanisms- have been equally cool.
Having fitted a number of experimental 3A maps, I can unequivocally say that the global structure is far from obvious
Okay, "obvious" was overstated. I guess I meant it relative to protein-folding simulations, which are beyond useless for telling you the tertiary structure. 3A is still enough to indicate the overall fold, i.e. "global structure", even if sidechains are incomprehensible.
CNS/CNX (the current versions of the molecular dynamics refinement program) works very well on multiprocessor systems, using OpenMP.
That's news to me - I've looked at the CNS source code and I couldn't find any parallel routines. I've seen stuff about using compiler auto-parallization, but this doesn't seem very effective. CNX may be parallel, but you have to shell out thousands of dollars to Accelrys for the priviledge of using it, and most academic labs would prefer to use the free version.
Regardless, I suspect the problems introduced by distributed computing would make it completely unfeasible for this type of work.
I am wondering whether he used distributed or parallel protien folding simulations for some of his work...
As the AC pointed out, none whatsoever. Almost nobody doing crystallography does, in fact. It continues to amaze me how many people think that simulations are going to replace experiment.
X-ray crystallography on globular protiens usually yields poor results (it is hard to get the X-rays to diffract to show the inner channel structure) compared to crystalline/regular protiens.
This is incorrect: crystallography of membrane proteins is extremely difficult compared to globular proteins. However, the techniques are getting better and better and there are now a number of very diverse membrane protein structures available.
Pumps are different. There is also a P-type ATPase, the calcium pump from muscle tissue, and the F1F0 ATPase, which pumps protons in the mitochondria. John Walker won the Nobel for the latter structures (although most of these were non-membrane). One of my former co-workers solved the cytochrome BC1 oxidase, also very large.
I think (can't remember for sure) that one reason the channels are so difficult is that they're nowhere near as stable when you take them out of the membrane. Intuitively, this makes sense comparing the structures: most of the others have coherent helical bundles, but the potassium channel does not. The key is allowing a very high rate of transport while maintaining excellent specificity.
I am willing to bet that it or other distributed computing projects are actually quite critical in the types of work represented by this Nobel prize.
Nope. Refinement of structures often uses molecular dynamics, one of the classical simulation methods and also a (very slow) way of looking at protein folding. However, the software that does this is single-processor and actually doesn't require too much more power than a fast desktop. These structures were all at around 3-Angstrom resolution, and once you're able to collect and properly phase X-ray data of that quality, the global structure is obvious.
Everything that was known or guessed about membrane channel structures before this work was done using electrophysiology and classical molecular biology - purely experimental methods, both quantitative and qualitative. Relatively few structural biologists use standalone simulations, and this is only ever done after a high-resolution structure has already been obtained. The only people using de novo simulations are those working on protein folding (most of whom are actually experimentalists) or protein design people.
I know of no important native biological structure solved with the aid of prior simulation; this is not to say that none exist, but you're grossly overestimating the importance of theoretical methods.
There was pretty much no doubt that MacKinnon would win it eventually - but it's a bit surprising that it came so soon, considering he's at the height of his career. He's only published four papers this year, but they're all Science or Nature (including one cover article). We can probably expect equally terrific work from him in the future.
I interviewed with him earlier this year (I applied to Rockefeller largely because of his lab), and he's one of the most intensely brilliant people I've ever met. There are very few scientists who will master a completely different technique in the middle of their career, while working on the same area of research. Fewer still are able to dominate the field. When I took physiology in college, we read multiple articles which described hypotheses proved by a single figure in one of MacKinnon's papers.
(There are actually an increasing number of membrane protein structures available, some of them quite large. However, ion channels are apparently especially difficult to study, and none were solved before MacKinnon started.)
It actually varies more than you'd expect. Kary Mullis didn't have to wait too long after PCR was invented, because the technique was immediately useful and revolutionary. Michel/Huber/Diesenhofer won in 1988 for X-ray structure of the photosynthetic reaction center, which they'd published only a few years previously. Until then, people had thought high-res structures of membrane proteins couldn't be determined.
Scientific research is weird, because what may in retrospect appear to be a "revolutionary" concept may take years to be recognized as such, even if they're immediately seen to be somewhat useful. And of course there are cases where a discovery was initially dismissed entirely, and the brilliance of those responsible for it wasn't realized until long after. Catalytic RNA and prions are the best recent example I can think of.
One of the places she worked at, Birkbeck College in London, suggests that she missed out on multiple prizes - after DNA, she worked on viral strucutres with Aaron Klug, who eventually won the Nobel himself.