Field of Science

Drugable.com "ranks billions of drug interactions"? Hold your horses.

Now here's a study that should make most seasoned molecular modelers cringe. Nature News reports on an effort by website Drugable.com that docked 600,000 compounds to 7,000 protein targets and predicted which ones would show activity against these targets based on docking scores:

Predicting how untested compounds will interact with proteins in the body, as Drugable attempts to do, is more challenging. In setting up the website, Cardozo’s group selected about 600,000 molecules from PubChem and the European Bioinformatics Institute’s ChEMBL, which together catalogue millions of publicly available compounds. The group evaluated how strongly these molecules would bind to 7,000 structural ‘pockets’ on human proteins also described in the databases. Computing giant Google awarded the researchers the equivalent of more than 100 million hours of processor time on its supercomputers for the mammoth effort.

But mammoth computing resources do not translate to carefully constructed protocols or correct predictions. In its current incarnation, docking is best for finding the binding pose, that is, the orientation of a drug bound into a protein pocket. Ranking compounds is far more difficult, and predicting absolute binding affinities is a very distant, currently unachievable third goal.

Anyone who has tried to run a hit to lead or lead optimization project based on docking scores would know how riddled with problems and qualifications any prediction based on these highly subjective numbers is. For starters, every modeling program gives you its own docking scores. Absolute values of these numbers (which ideally should reflect the free energy of binding but which seldom do) are almost always useless. If you are dealing with a congeneric series of molecules and are fairly confident about the binding orientation (usually confirmed by x-ray crystallography or some other technique) then maybe you could get some help from the scores in ranking the compounds, but even then mostly in terms of trends rather than quantitative differences.

Unfortunately the news piece says nothing about what method was used to generate the poses, whether there was any clustering or whether only the top pose was considered, what the false positive rate was, and most importantly, whether there was any experimental verification whatever of the ranking. The website is also not helpful in this regard. It also does not tell us if the protein structures used for docking were well-resolved or refined or whether they were homology models. In the absence of all this information the ranking of the compounds is tenuous at best and useless at worst and as it stands the study sounds little better than throwing darts in the dark and hoping some of them will stick. Ranking often fails even for similar compounds, so how well (or badly) it would work for 600,000 diverse compounds bound to 7,000 diverse protein targets is anyone's guess.

The report also compares the study to a similar activity prediction study by Brian Shoichet in which drug similarity was used to predict activity against unexpected targets. But that was a very different kettle of fish; it was a compound similarity - not docking - study so it did not have to deal with the complexities of error-ridden protein crystal structures or homology models, it verified a lot of the predictions using carefully constructed assays, and even then it gave a hit rate which did not exceed about 50%.

Either the Drugable.com study itself has failed to validate its predictions or the news report is woefully incomplete. Maybe I am wrong and in fact the study has laid the careful groundwork and validation that is necessary for trusting docking. As it stands however, the purpose of the report mainly seems to be to highlight the fact that Google generously donated 100 million hours of its computing power to the docking. This heightened, throw-technology-at-it sense of wonder and optimism is exactly what the field does not need. I would be the first one to welcome reliable predictions of drug-protein affinity based on wholesale docking of compounds to targets, but I don't think this work achieves that goal at all.

The future of nuclear energy: Let a thousand flowers bloom



The interior of a TRIGA nuclear reactor at Oregon State University (Image: Oregon State University)
In the summer of 1956, a handful of men gathered in a former little red schoolhouse in San Diego. These men were among the most imaginative scientists and engineers of their generation. There was their leader, Frederic de Hoffmann who had worked on the Manhattan Project and was now the president of the company General Atomics. Hoffmann was not only a creative physicist but also an unusually shrewd and capable manager and entrepreneur; in the later years of his life he would take the celebrated Salk Institute to great heights. There was also Freeman Dyson, a remarkably versatile mathematical physicist from the Institute for Advanced Study in Princeton who had previously reconciled disparate theories of quantum electrodynamics - the strange theory of light and matter. And there was Edward Teller, another Manhattan Project veteran; a dark, volatile and brilliant physicist who would become so convinced of the power of nuclear weapons to save the world that he would inspire the caricature of the mad scientist in Stanley Kubrick's classic film "Dr. Strangelove".

Together these men and their associates worked on a single goal: the creation of a nuclear reactor that was intrinsically safe, one that would cease and desist its nuclear transformations even in the face of human folly and stupidity. The reactor would have the rather uninspired name TRIGA (Training, Research, Isotopes, General Atomics) but its legacy would be anything but uninspiring. At the heart of the reactor's success was not a technical innovation but an open atmosphere of debate and discussion. Every day someone - mostly Teller - would come up with ten ideas, most of which sounded crazy. The others - mostly Dyson - would then patiently work through the ideas, discarding several of them, extracting the gems from the dross and giving them rigorous shape.

TRIGA benefited from a maximum of free inquiry and individual creativity and a minimum of bureaucratic interference. There was no overarching managerial body dictating the thoughts of the designers. Everyone was free to come up with any idea they thought of, and the job of the rest of the group was to either refine the idea and make it more rigorous and practical or discard it and move on to the next idea. The makers of TRIGA would have been right at home with the computer entrepreneurs of Silicon Valley a few decades later.

At the core of TRIGA's operation was a principle called the warm neutron principle. In a conventional reactor the neutrons in the fuel are moderated by hydrogen in the cooler water from the surroundings. There is a significant potential for a meltdown if someone pulls out the control rods, since the water which stays cool for a while will continue to moderate the neutrons and sustain their efficacy for causing fission. Dyson and Teller's idea was to place half of the hydrogen in the water and the rest in the fuel in the form of a uranium and zirconium hydride alloy. This would result in only half of the hydrogen staying cool enough to moderate the neutrons, while the other half in the hydride stays warm and diminishes the ability of the neutrons to fission uranium. This results in the fuel having what is called a negative temperature coefficient. The fuel rods were fashioned with care and precision by Massoud Simnad, an Iranian metallurgist working on the project.

The warm neutron principle is what made TRIGA intrinsically safe, very unlikely to sustain a meltdown or catastrophic failure. It took less than three years for the engineers and technicians to take the reactor from the design stage to manufacturing. The first TRIGA was inaugurated by none other than Niels Bohr in San Diego. Seventy of these safe reactors were built. They were safe and cheap enough to be operated in hospitals and universities by students and their main function was to produce isotopes for scientific and engineering experiments. They were also robust and safe enough to be proliferation resistant. As Dyson recounts in his elegant memoir "Disturbing the Universe", the TRIGA is perhaps the only nuclear reactor that made a profit for its creator.

TRIGA made the development of nuclear power seem relatively easy, cheap and fast. Why didn't other reactors enjoy the same success? Why, after fifty years, is nuclear power still struggling in the face of economics and political and social backlash? There are many reasons, but the principal reason is simple: the designers of TRIGA were encouraged to have fun and they had the kind of freedom of inquiry commonly found in a startup company. The problem is that the fun went out of the nuclear business in the 70s and with fun creativity and cost considerations also went out of the window. In his book Dyson swiftly cuts through to the central issue:
"The fundamental problem of the nuclear industry is not reactor safety, not waste disposal, not the dangers of nuclear proliferation, real though all these problems are. The fundamental problem of the industry is that nobody any longer has any fun building reactors....Sometime between 1960 and 1970 the fun went out of the business. The adventurers, the experimenters, the inventors, were driven out, and the accountants and managers took control. The accountants and managers decided that it was not cost effective to let bright people play with weird reactors. So the weird reactors disappeared and with them the chance of any radical improvement beyond our existing systems. We are left with a very small number of reactor types, each of them frozen into a huge bureaucratic organization, each of them in various ways technically unsatisfactory, each of them less safe than many possible alternative designs which have been discarded. Nobody builds reactors for fun anymore. The spirit of the little red schoolhouse is dead. That, in my opinion, is what went wrong with nuclear power."
Nobody builds reactors for fun anymore. What Dyson is getting at is quite simple. For any technological development to be possible, the technology needs to drive itself with the fuel of Darwinian innovation. It needs to generate all possible ideas - including the weird ones - and then fish out the best while ruthlessly weeding out the worst. This leads not only to quality but cost reduction since no entrepreneur is going to risk introducing an inherently expensive technology into the market. But all this is not possible until you allow people to play with ideas of their own volition and have fun doing it. People are not going to selflessly generate ideas by fiat, they are only going to do so when they are supported by funds and infrastructure but otherwise left to their own devices. The accountants and managers need to get the process started and then need to get out of the way.

Almost every successful technology has gone through this Darwinian phase. Dyson gives the example of motorcycles, which motorcyclists from his father's generation designed and serviced with care and affection. In our generation the most resounding example is that of computer technology. We have lost track of how many versions of software and hardware young computer enthusiasts experimented with in their California garages before their own technical and artistic sensibilities and the will of the market picked the best ones. Both Bill Gates and Steve Jobs made their fortunes in a milieu of young upstarts experimenting with the latest electronics and code and competing with fellow upstarts sprawled across the country. Just like the nuclear designers of the little red schoolhouse, the computer designers of the Silicon Valley garages were unencumbered by the demands of a central authority. So were the genetic engineers who founded companies like Genentech and Amgen. They could let their imaginations roam, bouncing ideas off one another and ruthlessly shooting down clumsy, expensive or ostentatious designs. It was the ability of bright young people to brainstorm to their hearts' content and to launch nimble startups rapidly exploring diverse and cheap technological solutions that allowed computer technology to become the all-pervasive life force that it is today. Biotechnology is now poised to do the same. A similar process of Darwinian survival of the fittest permeates other successful technologies, from flight to automobile engineering to house construction. And most importantly, the creators of all these technologies had fun creating them.

Nothing like this happened with nuclear power. It was a technology whose development was dictated by a few prominent government and military officials and large organizations and straitjacketed within narrow constraints. Most of the developers of nuclear technologies were staid, elderly bureaucrats rather than young iconoclasts like Frederic de Hoffmann. An early design invented by Admiral Hyman Rickover - suitable for submarines but hardly optimal for efficient land-based power stations - was frozen and applied to hundreds of reactors around the country. Since then there have been only a hundred or so reactor designs and only half a dozen or so prominent ones. Due to a complicated mix of factors including public paranoia, lack of economies of scale, political correctness and misunderstandings about radiation, nuclear technology was never given a chance to be played around with, to be entrusted to youthful entrepreneurs experimenting with ideas, to find its own way through the creative and destructive process of Darwinian evolution to a plateau of technological and economic efficiency. The result was that the field remained both scientifically narrow and expensive. Even today there are only a handful of companies building and operating most of the world's reactors.

To reinvigorate the promise of nuclear power to provide cheap energy to the world and combat climate change, the field needs to be infused with the same entrepreneurial spirit that pervaded the TRIGA design team and the Silicon Valley entrepreneurs. Young people who are brimming with ideas especially need to be given as many resources as possible to come up with solutions and explore them in startups, even if not garages. Just like any other technology, nuclear power can thrive only when the maximum number of people apply their creative minds to improving both the quality and cost of energy from fission. Fortunately a minority of companies and their creators are setting the trends.

I live in Cambridge, MA which has been a hotbed of innovation for several decades. In a few square miles along the picturesque Charles River lie literally hundreds of biotech, pharmaceutical and information technology startups, most enabled by the proximity of MIT and Harvard whose laboratories provide a steady supply of ideas that can be potentially turned into useful products. The scientists, engineers and managers in these startups constantly compete against each other and between themselves for the best ideas. My own startup is based on a novel way to make complex drugs using the specific base-pairing properties of DNA. Every year dozens of startups fail, and a few go public or are bought by other companies. The whole startup enterprise in Cambridge is subject to the forces of Darwinian selection that enables the filtering of the best ideas.

One component of this enterprise is named Transatomic Power. It was started by a duo of graduate students from MIT named Leslie Dewan and Mark Massie in 2010. The goal of Transatomic Power is to design a reactor that can generate power from nuclear waste, thus addressing the twin issues of clean energy and nuclear waste removal at the same time. The reactor which is a molten salt reactor lives off the preponderance of energy trapped in unfissioned reactor fuel from light water reactors. It is also compact enough to be shipped individually to the reactor site. Dewan and Massie are two of the few young people who actually see opportunity in the nuclear field and are willing to take risks in order to develop a novel approach to the problem.

On the other coast of the United States in Seattle is another team of nuclear entrepreneurs led by Nathan Myhrvold, a former CTO of Microsoft with degrees in physics and economics. Myhrvold has founded a company named Terrapower which operates on a novel nuclear design called the traveling wave reactor (TWR) which was also in part explored by Edward Teller and Lowell Wood in the 90s. The TWR is another reactor which can operate on waste, using depleted uranium to sustain a fission wave that spreads outward into the reactor, transforming the uranium into plutonium and leaving a small amount of fissile waste behind. The TWR promises to run for decades without having to refuel it or recover spent fuel, thus promising both safety and proliferation resistance. Among the enthusiasts of the TWR is Bill Gates, who knows a thing or two about Darwinian innovation in technology.

The founders of Terrapower and Transatomic are following in the footsteps of the dreamers in the little red schoolhouse. They have transformed nuclear technology into an entrepreneurial game of ideas and funding sustained by a healthy interplay between academic, industrial and government laboratories. I do not know whether their reactors will be the ones supplying the world's energy in the near future, but what I do know is that they are doing exactly what needs to be done to sustain the innovative process of creation and destruction that is necessary for the evolution of any successful technology. They are bucking the trend set by the large, bureaucratic government organizations and their industrial counterparts. And most importantly, they are having fun doing it, trading ideas and exploring new technical ground. I see hope in the adventures of these nuclear explorers, just like the makers of TRIGA saw hope in the future of nuclear power and the whole world saw hope in the explorers of computer and biotechnology in the 80s. When it comes to nuclear technology we should let a thousand flowers bloom. And then we can pick the most beautiful.

This post was first published on the Nobel Week Dialogue website.