Ironically, I don't think it's possible to beat chemical-based energy sources for providing heat.

Efficiency is nearly ideal when you want energy and not work, so I can easily envision electric vehicles carrying a supplementary propane system to provide cabin (and battery) heating in the cold. There are also propane-fueled AC units, so it could serve double duty. The simplicity and low weight of such an auxiliary system makes it little more than an afterthought addition to the overall engineering.

I won't argue that many new cars are unnecessarily convoluted and make use of proprietary systems. However, a lot of the basic vehicle maintenance and repair issues aren't assisted much by computer diagnostics: in general they are only useful for reporting on sensor status.

I have an example of "backyard mechanic-ing" on a modern vehicle. A friend's Audi A4 was misfiring, and the OBDII system wasn't really reporting anything useful. This isn't surprising, since there isn't any feedback on the ignition circuits, which include one coil per cylinder, and is fully electronic (no distributor, etc).

I performed diagnostics with some LEDs ripped out of a scrap computer case I has lying around. Tapping into the various control circuits with the LEDs functioning as noid lights let us see the ignition pulses through the various components of the system. In the end it turned out to be a problem with a glorified, overpriced relay.

Computer diagnostics would have been no use to us. To my knowledge, this type of scenario (substantial engine components without feedback circuits) is the standard, and I've run into it often enough with vehicles to plan an Arduino-based monitoring and diagnostic system.

I'm guessing that you haven't been involved with actual NIH/NSF funding.

Resource allocation isn't performed randomly: it mirrors actual real-world concerns. Cancer research is huge, as is Alzheimer's. Energy-related research has jumped in popularity in recent years.

The problem with a private-sector approach to research is that by definition it's going to involve secrecy until you find something marketable. There would be large incentives to keep every discovery private. With the current system, you only need to keep your discoveries private until they can be published.

That's a much lower barrier to information-sharing, and the end result is that there is much more communication going on, which helps everyone and discourages resource-wasting duplicated work. In addition, the publication barrier is dropping every year, with more credence given to pre-publication resources as a way to stake your claim in a field.

If it comes down to selfish profiteering vs altruistic sharing, sharing is a better way to build knowledge.

You said that the patent "covers the procedure of looking at the results", which implies that a procedure was specified and patented.

That's not the case: they did not patent any procedure. Their patent is more or less covering the whole concept of these compounds in a medicinal context. Any analysis at all would be infringing.

Perhaps I was nit-picking too much.

No, they patented the very idea of looking at those chemicals, as you recognized later in your statements. That's exactly why there was no good reason to issue these patents.

Well, if you're looking for the application of knowledge, the patents are sorely lacking. They contain sentences like: "The level of 6-TG can be determined, for exacmple, in red blood cells using high pressure liquid chromatography (HPLC)". The same sentence is repeated elsewhere with the name of a different chemical substituted in for 6-TG.

A really novel HPLC method might be patentable, but that's not part of their patent at all. They just want to patent the idea of looking at these chemicals as a diagnostic tool.

Why should that be revolting?

I was discussing general standards of protocols, as carried out by typical researchers. The issue of shade-tree R&D vs professional was not part of the equation at all. Good results, bad results... that's a totally unconnected tangent.

I'm talking about cost/benefit analyses, and the positives and negatives of a particular approach. Everyone should make that kind of evaluation regularly, and even the unexpected can be considered: what's the possibility of making a discovery that has been overlooked?. For example the cost and effort of mixing a pack of chesterfield kinds with water and soap is nearly nothing, the potential negatives (killing all the plants?) are pretty small, and benefits could be substantial (cost-effective non-toxic pesticide?). So it would make perfect sense to try such a thing, if you wanted to.

Well, you've succeeding in taking a (mostly) accurate statement used to make a general point, and drug it through several layers of unnecessary quibbling. Congratulations.

The whole point of my original statement was to illustrate that outsourcing is now both commonplace and necessary. Sequencing now costs $5 per sample, and will get you 800+ bp. That cannot be matched by non-specialists.

Just be honest about it: what kind of quality level could the average researcher achieve, at what cost, assuming that a sequencer was available? I'm sure we could all learn how to do it properly given enough time, but here are my estimates: for 95+% of the general researcher population, a 99% confidence/success rate would be maybe 50-100 bp maximum, and the costs and time would be more than 10-fold greater than outsourcing. That is laughable and pathetic compared to current standards.

You have the expertise and equipment to dump a Kraft bottle of Thousand Island over an uncut head of iceberg lettuce. That doesn't mean you can make a decent salad.

It's true that it's possible to accomplish a great deal of biology/biochemistry research using just basic tools: I would say that the single greatest analytical tool in biochemistry is the polyacrylamide gel, which can be produced and used with no real specialized training or tools.

However, we're moving away from such "crude" techniques towards more sophisticated analytical tools, since in many ways biochemistry is now technology-limited. Single-molecule work, such as that pioneered by Carlos Bustamante provide insights that would never be possible with classical methods, and on the other end of the spectrum, we're now working on characterizing the entire network of small metabolite molecules simultaneously and quantitatively. This kind of work just isn't easily carried out by amateur enthusiasts.

That said, there is certainly quite a bit of research that DIY biologists would be capable of performing, especially considering that they could have access to the same kind of resources that professionals do. For example, after amplifying a gene, no researcher will sequence it themselves: it's shipped of to a specialized lab that will do it, for a fee. That sequencing step requires equipment and expertise that's at a higher level than even the pros don't have.

But regardless of theoretical ability, the professionals retain the advantage that it is their job to work on these projects. The time they can dedicate to their work will be far greater than someone who does it as a hobby.

Back to the subject of "openness", the professional scientific world isn't nearly as closed-off as the article would have you believe. It is true that there is a persistent fear of being "scooped", but the standards are changing for staking your claim on a particular piece of research.

It used to be that a full manuscript in a scientific journal was the only thing sufficient to get credit for something. Now, people are gradually embracing online resources are a valid way to communicate, and by extension, to prove that they were the source of any particular bit of publicized material. Even non-finalized material is now more common to make public: Nature has a pre-publication online source for publishing findings, and there are journals devoted entirely to negative results, which was previously unheard-of.

The walls are coming down, it's just a question of finalizing the transition, and winning over the old guard.

Disclosure: I am a professional research scientist, one of the younger ones. I have a substantial hardware/software project in the works, which will likely be simultaneously published via classic journal, online website, and software via SourceForge.

In the numbers you posted above, you didn't include vacated decisions. Based on my limited understanding of law, I think it's reasonable to include those in the "wrong decision" category. If you include the vacated decisions, the ratio of poor decisions goes up, as expected.

http://spreadsheets.google.com/ccc?key=rOMUZthYouib7bNYZMyw5rg

But regardless of whether you include the vacates, the distribution of "wrong decisions" is clearly different between the 9th and the others. I think a Kolmogorov-Smirnov test might be appropriate, and if you apply that, the conclusion is that that Ninth Circuit is absolutely different from the rest.

So why is the Ninth different? When dealing with affirmed/(vacated+reversed) ratios, size doesn't matter, except that small sample sizes increase the noise: that's the reason I grouped the K-S test into two categories, Ninth and !Ninth.

The other bias I can think of is that there is some geographic oddity of the Ninth's location, which produces inexplicable biases. Seems unlikely to me.

So what is left besides either judicial bias, or workload problems causing poor ruling?

It would really take an expert in DNA folding (such as the authors of the paper) to give you a good answer to that.

But here's my partially-educated guess as to why DNA folds "better": there are very few examples in which the very first folding steps for a protein is understood. As of a year or two ago, it was still up for debate which kind of interactions were the most important ones for forming the intial "seeds" that would lead to a fully-folded structure. Without being able to control the start of the folding, the search space for a random configuration to find the correct final fold is unimaginably huge.

In contrast, DNA folding follows more simple rules, and the initial folding steps can be easily controlled. So assuming you can initialize correct folding by properly engineered sequences, you just have to make sure it continues along the path. That makes it a directed, and much simpler, problem.


The stability of a DNA structure vs protein is going to depend highly on the specifics. But, you can design a double-stranded DNA segment that will separate into two individual strands at a very precise temperature, because you can specifically control the number of bonds (in a particular segment). It doesn't take a lot to get stability into the 80-100degC range, but that's just for two strands together, not for a full cage. I'm not sure at what point you would lose that level of stability.


For proteins, stability ranges across the whole spectrum. Some nanostructures fall apart if the salt concentration is just a little off, while others will be just fine near boiling: there are viruses that survive great in the geothermal features in Yellowstone.

From a theoretical perspective, there are many reasons why a "stapled" DNA structure would be preferred to more convoluted one-piece structure.

Think of it as a modular structure: the individual components give you flexibility in tuning a structure to fulfill a variety of roles. The cage could be fine-tuned to assemble or disassemble at particular rates, or with variations in size. Each "staple" location is a site where you can add a modification to give new functionality. For example, the display of some short peptide sequences are sufficient to cause an organism to traffic the particle to one type of tissue exclusively. Modularity is a good thing.


As for practicality, perhaps you should consider that every cell in your body employs nanostructures as part of its general operation. As nanoscale engineers we're far behind the curve at even replicating what is currently in existence. There are thousands of problems out there just waiting to be solved as soon as we figure out how to do it.

As a biochemist working in the area of structure/physics, of course I find this very interesting, and there's no shortage of things that could be said about this technique.

However, one of the most relevant issues in biotech and nanotech is the question of cost. The most elegant drug delivery system in the world will never be viable if you can't produce it in decent yields, at a reasonable cost.

My work involves viral capsids, which we use as nano building blocks because they (sometimes) self-assemble, making very large, symmetric structures with relative ease. However, you still have to produce the protein, which usually involves engineering some other organism to produce it for you, since it can't be done synthetically. Assuming that step can be accomplished, you still must purify it, and hope that once all is said and done the protein has retained the appropriate structure. If it's been "deformed" along the way, it's usually a one-way street, and your precious product is now garbage.

In contrast, DNA can be made more or less fully synthetically, and the misfolding problem is a non-issue: it can be melted down and re-folded nearly infinitely.
Those features make DNA really interesting as a better candidate for commercially-viable nanotech. On the other hand, DNA is going to be uniformly negatively charged everywhere, as opposed to proteins which can take on nearly any characteristic you might want, due to the range of amino acid building blocks. In a biological sense such as the article mentions, that could be a concern if you want it to interact with (or avoid) other structures.

It should be possible. Other comments here (http://science.slashdot.org/comments.pl?sid=1221551&cid=27821391) have indicated that the bottle should "off-gass", and those gases could be collected for analysis. Sensitivity might be a concern, since the best gas-sampling mass spectrometers aren't usually the best at measuring isotope ratios, which is required for the analysis.

Yes, sorry, I should have included that caveat.
From the CDC:
* Heart disease: 652,091
* Cancer: 559,312
* Stroke (cerebrovascular diseases): 143,579
* Chronic lower respiratory diseases: 130,933
* Accidents (unintentional injuries): 117,809
* Diabetes: 75,119
* Alzheimer's disease: 71,599
* Influenza/Pneumonia: 63,001
* Nephritis, nephrotic syndrome, and nephrosis: 43,901
* Septicemia: 34,136

After looking up the numbers, I must admit that it's closer that I was expecting: perhaps 870k for health-related issues vs 270k for infectious agents. With the current trends, I think we'll see that gap increase.