Really old news. MS has been doing this sort of thing for years, the old "your company is underlicensed" trick. "Would you believe you need 500 licenses?" These stories have been reported before, and IT managers have been insprired to switch to OSS as a result, although not very many.
that's not really accurate. TB and malaria kill millions each year and neither are viruses. there are many types of bad bugs. viruses are bad, but they are not the only thing to worry about.
they have been combining antibiotics for decades. it is absolutely necessary to treat things like TB and gonorrhea these days, whereas before 1 drug would do the trick nicely.
the problem is that the bugs will not be eradicated. they will continue to divide, and each division brings about the chance for resistance. also, the bugs "in the wild" will be exposed to each drug individually, allowing for development of resistance to each drug. this resistance can and will be moved "horizontally" across species barriers by transposable elements.
a lot of this comes from careless use of these drugs and poor patient compliance.
for this particular type of movement, maybe so. but the point is to develop the ability to have a robot-biological organism interface that might ultimately allow for far more complicated control scenarios. for example, the brain of a small amphibian would likey be much better at controlling a vehicle navigating varied terrain than the massive computers currently stuffed into "automated" robotic cars.
of course, we may need to worry about the slime mold becoming self aware, forming SlimeNet, and then destroying humanity.
Indeed, there are techniques on the horizon to study atomic resolution without crystals, and the ability to observe a single molecule may be someday routine. But for atomic resolution (~1 angstrom or so) at ths time, nothing beats Roentgen's x-ray, equations of Laue and the Bragg, Bernal and Crowfoot's crystallization techniques, and an obscenely powerful digital computer (even though proteins were crystallized in 1918, the first structure coincides with digital computers 40 years later, as 3-D fourier transforms on this scale are not feasible using any other method).
NMR is not really "more useful", and certainly not because it is "in solution". The distortions of the NMR model are likely to be greater than those from the crysallographic model. A lot of work has gone into studying the caveats of each technique, and it is not meant to be a pissing contest. They are both tools, and each has a job. But people (including biochemists) hold the misconception that NMR is easier than crystallography, and more accurate. neither is remotely close to true.
Not sure what you mean by "non-native" solution. The protein will (hopefully) be in a native conformation in the buffer. But NMR does suffer from size limitations and is only possible for rather small proteins, at the moment. Advances in pulse sequences may improve the performance. (Nuclear magnetic resonance relies on radio pulses in a very strong magnetic field (>10 Tesla) to observe "relaxation" in the aligned molecules. It is the same as MRI, but no one wants a medical procedure with "Nuclear" in the title.)
Also, the solution to NMR data may have quite a bit more error in it than crystallography. NMR is not a "global" fit, but rather a mathematic solution relating the positions of each atom to its nearest neighbor, so which each nearest neighbor you can propagate a small error that can get large across the whole structure. This is usually more of a problem for nucleic acids than proteins, and the use of dipolar couplings can improve the global fit of the model.
At this point, x-ray crystallography is more for analyzing complexes, which tend to be larger, and NMR more for fragments, but they each have their niche.
as a final note, just because NMR does not rely on crystals does not mean it is easy. It is very difficult to find a buffer condition that keeps the protein at extremely high concentration yet does not precipitate, etc. And that does not even get into the problems surrounding collecting clean data.
Slashdot Mirror
Really old news. MS has been doing this sort of thing for years, the old "your company is underlicensed" trick. "Would you believe you need 500 licenses?" These stories have been reported before, and IT managers have been insprired to switch to OSS as a result, although not very many.
that's not really accurate. TB and malaria kill millions each year and neither are viruses. there are many types of bad bugs. viruses are bad, but they are not the only thing to worry about.
they have been combining antibiotics for decades. it is absolutely necessary to treat things like TB and gonorrhea these days, whereas before 1 drug would do the trick nicely.
the problem is that the bugs will not be eradicated. they will continue to divide, and each division brings about the chance for resistance. also, the bugs "in the wild" will be exposed to each drug individually, allowing for development of resistance to each drug. this resistance can and will be moved "horizontally" across species barriers by transposable elements.
a lot of this comes from careless use of these drugs and poor patient compliance.
for this particular type of movement, maybe so. but the point is to develop the ability to have a robot-biological organism interface that might ultimately allow for far more complicated control scenarios. for example, the brain of a small amphibian would likey be much better at controlling a vehicle navigating varied terrain than the massive computers currently stuffed into "automated" robotic cars.
of course, we may need to worry about the slime mold becoming self aware, forming SlimeNet, and then destroying humanity.
Indeed, there are techniques on the horizon to study atomic resolution without crystals, and the ability to observe a single molecule may be someday routine. But for atomic resolution (~1 angstrom or so) at ths time, nothing beats Roentgen's x-ray, equations of Laue and the Bragg, Bernal and Crowfoot's crystallization techniques, and an obscenely powerful digital computer (even though proteins were crystallized in 1918, the first structure coincides with digital computers 40 years later, as 3-D fourier transforms on this scale are not feasible using any other method).
NMR is not really "more useful", and certainly not because it is "in solution". The distortions of the NMR model are likely to be greater than those from the crysallographic model. A lot of work has gone into studying the caveats of each technique, and it is not meant to be a pissing contest. They are both tools, and each has a job. But people (including biochemists) hold the misconception that NMR is easier than crystallography, and more accurate. neither is remotely close to true.
Not sure what you mean by "non-native" solution. The protein will (hopefully) be in a native conformation in the buffer. But NMR does suffer from size limitations and is only possible for rather small proteins, at the moment. Advances in pulse sequences may improve the performance. (Nuclear magnetic resonance relies on radio pulses in a very strong magnetic field (>10 Tesla) to observe "relaxation" in the aligned molecules. It is the same as MRI, but no one wants a medical procedure with "Nuclear" in the title.)
Also, the solution to NMR data may have quite a bit more error in it than crystallography. NMR is not a "global" fit, but rather a mathematic solution relating the positions of each atom to its nearest neighbor, so which each nearest neighbor you can propagate a small error that can get large across the whole structure. This is usually more of a problem for nucleic acids than proteins, and the use of dipolar couplings can improve the global fit of the model.
At this point, x-ray crystallography is more for analyzing complexes, which tend to be larger, and NMR more for fragments, but they each have their niche.
as a final note, just because NMR does not rely on crystals does not mean it is easy. It is very difficult to find a buffer condition that keeps the protein at extremely high concentration yet does not precipitate, etc. And that does not even get into the problems surrounding collecting clean data.