Not only are most proteins not able to use this method, protein crystallography is still as much an art as a science. The process of forming a protein crystal requires a protein solution that very slowly becomes super-saturated to the point that the protein molecules start to clump together. To be any good for crystallography, that clumping has to be very controlled (slighly negative second virial coefficient.) If the clumping is too rapid or too favorable, the protein will just crash out like a scrambled egg. If its too slow or repulsive you can wait years to get a decent crystal.
As it has been pointed out lysozyme is NOT a good test protein -- its been done way to often and will crystallize in dozens of different conditions. Real interesting proteins are difficult to crystalize and sometimes will crystalize once and cannot be repeated. This has led to high-throughput robotic methods of trying thousands of different crystallization conditions. Investing the time and effort to perfect a single condition with magnetic levitation may only be useful once small crystals can be developed by other methods.
I think that the real benefit of this work will be to expand the use of neutron diffraction. Current neutron sources are far less intense than x-ray sources. So increasing the crystal size can help a lot. Finally, it might seem counterintuitive but sometimes "perfect" crystals are actually bad -- too small of a crystal mosaicity (how disordered a crystal is) can actually make getting good diffraction data difficult. In these cases (which unfortunately I have not experienced) the crystals are actually purposefully deformed in order to get measurable diffraction data.
All that being said, anyone know where I can get a 33 tesla magnet? Besides having fun levitating small amphibians, I have this paramagnetic protein I'd like to make into jewlery. . .
While I would love to live in your world where people's and companies priorities are right, I don't believe that either you or I live in that world.
The university department where I work now probably spends more money on administration than on teaching and research. Think of how much time and effort politicians spend on getting elected and how little it seems they spend on getting their promises turned into law. Also, I know how much time I wasted in college playing games and goofing off instead of studying.
These are not all evil -- without the administrative structure universities would collapse into a chaotic mess. Without spending millions of dollars on elections, politicians wouldn't be able to attempt to promote their (sometimes) good ideas. Finally, without goofing off, I honestly don't think I would have made it through college.
Pharmaceutical companies are in business. If they have to spend millions on advertising to make money to spend on research, I'm fine with that (to a point). As long as the companies don't lie, cheat, or steal, they should be allowed to do what every other company, group, and person does.
I've used this system in a practice setting along with a few other medical information systems -- the VA system is by far the most complete and useful. This sentiment is repeated by all of the doctors and medical professionals that I've worked with. It has features that allow for easy tracking of prescriptions, allergies, and known diagnoses which decreases the chance for medical errors.
The main impediment to wide adoption is that since its so complete, its can be complex to use, let alone administer. However, after about 1 week using the system in a real medical practice, a practictioner can become familiar enough to dread ever going back to another computer system (or god forbid, a paper chart).
The way that Unisys provides 99.9 percent uptime is by migrating processes from one processor to another when Windows crashes. The newest systems are able to run multiple OSes on the same machine (including multiple copies of Win). So when an NT crash happens, the entire system is not taken down and data can be salvaged. In that case, how do you measure uptime -- the machine (hardware) could be run for years, shutting down processors and replacing them when needed without turning off the server?
Slashdot Mirror
Not only are most proteins not able to use this method, protein crystallography is still as much an art as a science. The process of forming a protein crystal requires a protein solution that very slowly becomes super-saturated to the point that the protein molecules start to clump together. To be any good for crystallography, that clumping has to be very controlled (slighly negative second virial coefficient.) If the clumping is too rapid or too favorable, the protein will just crash out like a scrambled egg. If its too slow or repulsive you can wait years to get a decent crystal.
As it has been pointed out lysozyme is NOT a good test protein -- its been done way to often and will crystallize in dozens of different conditions. Real interesting proteins are difficult to crystalize and sometimes will crystalize once and cannot be repeated. This has led to high-throughput robotic methods of trying thousands of different crystallization conditions. Investing the time and effort to perfect a single condition with magnetic levitation may only be useful once small crystals can be developed by other methods.
I think that the real benefit of this work will be to expand the use of neutron diffraction. Current neutron sources are far less intense than x-ray sources. So increasing the crystal size can help a lot. Finally, it might seem counterintuitive but sometimes "perfect" crystals are actually bad -- too small of a crystal mosaicity (how disordered a crystal is) can actually make getting good diffraction data difficult. In these cases (which unfortunately I have not experienced) the crystals are actually purposefully deformed in order to get measurable diffraction data.
All that being said, anyone know where I can get a 33 tesla magnet? Besides having fun levitating small amphibians, I have this paramagnetic protein I'd like to make into jewlery. . .
While I would love to live in your world where people's and companies priorities are right, I don't believe that either you or I live in that world.
The university department where I work now probably spends more money on administration than on teaching and research. Think of how much time and effort politicians spend on getting elected and how little it seems they spend on getting their promises turned into law. Also, I know how much time I wasted in college playing games and goofing off instead of studying.
These are not all evil -- without the administrative structure universities would collapse into a chaotic mess. Without spending millions of dollars on elections, politicians wouldn't be able to attempt to promote their (sometimes) good ideas. Finally, without goofing off, I honestly don't think I would have made it through college.
Pharmaceutical companies are in business. If they have to spend millions on advertising to make money to spend on research, I'm fine with that (to a point). As long as the companies don't lie, cheat, or steal, they should be allowed to do what every other company, group, and person does.
I've used this system in a practice setting along with a few other medical information systems -- the VA system is by far the most complete and useful. This sentiment is repeated by all of the doctors and medical professionals that I've worked with. It has features that allow for easy tracking of prescriptions, allergies, and known diagnoses which decreases the chance for medical errors.
The main impediment to wide adoption is that since its so complete, its can be complex to use, let alone administer. However, after about 1 week using the system in a real medical practice, a practictioner can become familiar enough to dread ever going back to another computer system (or god forbid, a paper chart).
The way that Unisys provides 99.9 percent uptime is by migrating processes from one processor to another when Windows crashes. The newest systems are able to run multiple OSes on the same machine (including multiple copies of Win). So when an NT crash happens, the entire system is not taken down and data can be salvaged. In that case, how do you measure uptime -- the machine (hardware) could be run for years, shutting down processors and replacing them when needed without turning off the server?