The same group of researchers published a paper in 2009 in the journal Science using a technique called atomic force microscopy (AFM) rather than the scanning tunneling microscopy (STM) approached used here. This technique allowed them to resolve the atomic structure of pentacene, showing the classic ring structure as one might see drawn on a chalk board in their chemistry class. Combined with their means of imaging molecular orbitals by STM, these researchers have developed some really nice tools for studying molecules. Here's the citation for the AFM paper:

Gross et al. (2009) The Chemical Structure of a Molecule Resolved by Atomic Force Microscopy. Science, 325: 1110. doi:10.1126/science.1176210

Well, let's say these engineered worms escape into the environment. 1) the paper does not show whether the changes they made to the worm's genome are heritable, so the worm's offspring might not be able to incorporate the unnatural amino acids and the trait might go away after the escaped engineered worms die. Even if the trait is heritable, the paper suggests that the gene cassette they engineered into the worm gets lost from the genome over time, so after a few generations, the trait would likely be lost. 2) these worms do not have the ability to synthesize the unnatural amino acids on their own. They incorporate the unnatural amino acids into their proteins only when the researchers feed the worms large amounts of the unnatural amino acid. Without a source of unnatural amino acids, they are just slightly broken versions of a normal C. elegans worm.

Does this make you feel any better?

In the case of this study, the researchers are tricking the worm to incorporate an unnatural amino acid in the place of a stop codon (TAG to be specific). The researchers created an reporter gene that codes for a red fluorescent protein (mCherry) after a TAG stop codon. If the worm is not able to incorporate the unnatural amino acid, the cell will stop producing the protein once it encounters the TAG stop codon and not produce the red fluorescent part of the reporter gene. Successful incorporation of the unnatural amino acid, however, allows the cell to bypass the TAG stop codon and produce the red fluorescent part of the reporter gene. So, even though the unnatural amino acid is not directly producing the fluorescence, seeing the worms glowing red means that the worm's cells were able to incorporate the unnatural amino acid successfully.

You actually wouldn't need to even start with Vaccinia virus. In 2008, scientists at the J Craig Venter Institute synthesized and assembled the genome of an entire bacterium from scratch (Gibson et al. 2008. Complete Chemical Synthesis, Assembly, and Cloning of a Mycoplasma genitalium Genome. Science 319: 1215 - 1220. doi:10.1126/science.1151721) . The bacterial genome they synthesized was 580,000 base pairs compared to the 186,000 base pair size of the smallpox genome. Of course, commercial gene synthesis companies would never sell anyone any sequence resembling a smallpox sequence, but given enough resources, a government or even some well funded group could conceivably resurrect smallpox without needing a sample of the virus.

With new gene synthesis technologies (such as the ones used by Craig Venter to create his "synthetic" bacteria), one could fairly easily resurrect smallpox from DNA chemically synthesized in the lab along with some engineered cell lines. This was demonstrated with poliovirus nearly a decade ago (Cello, Paul, and Wimmer 2002 Chemical Synthesis of Poliovirus cDNA: Generation of Infectious Virus in the Absence of Natural Template Science 297: 1016. doi:10.1126/science.1072266). So, even eliminating every remaining sample of smallpox on Earth would not guarantee that the virus could not one day be resurrected as a bio-weapon.

Here's a small detail that the article leaves for the last paragraph:

In other words, they can zip DNA through this device quickly and measure some signal as the DNA passes through, but no one knows yet whether it is possible to extract accurate sequence information from the signal they get. Similar implementations (that admittedly have a less sensitive way of getting a signal from the different DNA bases) have so far failed to see significant enough differences between the DNA bases to be useful for sequencing. It's not clear that this method will work as advertised

It's typical of mathematicians. They construct a well posed problem and convincingly show that a solution exists, but make no effort at finding the actual solution to the problem.

There are existing techniques that give ~tens of nanometers resolution using fluorescence microscopy (discussed in a feature in Nature Methods ). Techniques such as PALM/FPALM/STORM (developed by Betzig, Hess, and Zhuang, independently) use photoswitchable fluorophores to image and localize single fluorescent molecules with high precision then reconstruct the image from these single molecule images. Another technique, STED (stimulated emission depletion, developed by Hell) uses stimulated emission to effectively shrink the size of the point spread function of a fluorescence microscope. Yet another technique, structured illumination microscopy (developed by Gustafsson), plays tricks with moiré patterns to extend the resolution of optical microscopy. All would, in theory, be applicable on Smith's array tomography samples.

On issue with superresolution fluorescence microscopy, however, is that the spatial resolution of an image is dependent on the density of antibodies bound to the sample. The Nyquist criterion defines how frequently one must sample the underlying structure (the neuron) in order to achieve a specific spatial resolution. In this case, each antibody that binds to the neuron is one sampling event. Therefore, achieving very high resolution requires binding more antibody to the sample than typical for standard immunohistochemistry. This can be difficult, especially in samples that are embeded in resin (as is required to get the 70 nm sections used in the array tomography method), as the embeding process can drastically reduce the antigenicity of the sample.

You are correct that the site did not correctly format the DOI link, but the research has been published. Here is the correct DOI link doi:10.1038/nature09518. Also, here is the link to the article on the Nature website: http://www.nature.com/nature/journal/vaop/ncurrent/abs/nature09518.html (link probably valid only for the next week or so)

where they leave a carefully hidden note that says they basically just copied the article verbatim from a university press release (e.g. see Science Daily and other related sites).

Thank you two for the clarification. I should really read the links more thoroughly before posting.

I don't know where that 34% figure comes from for the manual double checking. The test set contains about 60% vandalism and 40% real edits, so I'll assume this represents the rate of vandalism on wikipedia. Now, consider a set of 1000 edits. 600 would be vandalism while 400 would be real edits. The second filter would catch 570 instances of real vandalism along with 120 false positives. Even if you used the first filter to automatically remove the 120 instances of vandalism it finds, you would still be left with a set of 450 instances of vandalism + 120 false positives to check. This means that you would have to sort through about 57% of the original edits in order to identify the 120 false positives.

These are all good points, especially the point that cryo-EM helps avoid artifacts induced by crystal packing. However, I'm not sure cryo-EM helps with the issue of conformational flexibility of these large complexes. Even though the cryo-EM samples will freeze the sample in whatever conformation it was in, the single particle reconstruction methods used to analyze the EM data will average over the conformational heterogeneity, so you would end up losing most of that information in your final structure anyway. The same averaging over conformational heterogeneity occurs in crystallography except in the data acquisition step instead of the data analysis step. While it is in principle possible to sort out the different conformers in your sample during image classification, this is very difficult to do (because the cryo-EM images have such poor resolution and low contrast) and would likely only catch very large structural rearrangements.

The Aug 27 issue of Science, in which the x-ray crystallography study from Scripps appears, actually pubished two papers that describe the structure of adenovirus. The two papers use different techniques to achieve the same ends: the study from the researchers at Scripps grew crystals of the virus and studdied the x-ray diffraction patterns to deduce the structure of the virus. The other paper, done in collaboration between researches at UCLA and Xiangtan University in China, used a technique called cryo-electron microscopy. In this technique, the researchers freeze samples of the virus and use an electron microscope to take tens of thousands of pictures of different viruses within their samples. Although the pictures only give 2D projections of the virus structure, the individual electron microscopy images show the virus from different perspectives. By computationally aligning the images, they can reconstruct the 3 dimensional structure of the virus from the many 2D images taken. While this technique avoids the inherent difficulties of producing crystals (a process that can take decades for some samples), until very recently it has been difficult to achieve high resolution structures using this method. The cryo-EM adenovirus structure is one of only a handful of atomic resolution cryo-EM structures that have been solved to date.

While both studies are very informative and represent scientific tours de force for their respective techniques, it is interesting that the Medical Daily focuses only on the x-ray crystallography study from Scripps. Indeed, in a commentary published by Science that accompanies the articles, Prof. Stephen Harrison of Harvard Medical School (the first person to describe the full structure of a virus) writes that, "Indeed, the cryo-EM density map of Liu et al. appears to be substantially clearer and more interpretable than the x-ray density map of Reddy et al." Perhaps Medical Daily needs to do a better job of doing their homework.

The cryo-EM study is available at the following link (subscription required): http://www.sciencemag.org/cgi/content/abstract/329/5995/1038