Computer science: The learning machines

Using massive amounts of data to recognize photos and speech, deep-learning computers are taking a big step towards true artificial intelligence.

Article tools
PDF
Rights & Permissions

BRUCE ROLFF/SHUTTERSTOCK

Three years ago, researchers at the secretive Google X lab in Mountain View, California, extracted some 10 million still images from YouTube videos and fed them into Google Brain — a network of 1,000 computers programmed to soak up the world much as a human toddler does. After three days looking for recurring patterns, Google Brain decided, all on its own, that there were certain repeating categories it could identify: human faces, human bodies and … cats1.

Google Brain's discovery that the Internet is full of cat videos provoked a flurry of jokes from journalists. But it was also a landmark in the resurgence of deep learning: a three-decade-old technique in which massive amounts of data and processing power help computers to crack messy problems that humans solve almost intuitively, from recognizing faces to understanding language.

Deep learning itself is a revival of an even older idea for computing: neural networks. These systems, loosely inspired by the densely interconnected neurons of the brain, mimic human learning by changing the strength of simulated neural connections on the basis of experience. Google Brain, with about 1 million simulated neurons and 1 billion simulated connections, was ten times larger than any deep neural network before it. Project founder Andrew Ng, now director of the Artificial Intelligence Laboratory at Stanford University in California, has gone on to make deep-learning systems ten times larger again.

Such advances make for exciting times in artificial intelligence (AI) — the often-frustrating attempt to get computers to think like humans. In the past few years, companies such as Google, Apple and IBM have been aggressively snapping up start-up companies and researchers with deep-learning expertise. For everyday consumers, the results include software better able to sort through photos, understand spoken commands and translate text from foreign languages. For scientists and industry, deep-learning computers can search for potential drug candidates, map real neural networks in the brain or predict the functions of proteins.

“AI has gone from failure to failure, with bits of progress. This could be another leapfrog,” says Yann LeCun, director of the Center for Data Science at New York University and a deep-learning pioneer.

“Over the next few years we'll see a feeding frenzy. Lots of people will jump on the deep-learning bandwagon,” agrees Jitendra Malik, who studies computer image recognition at the University of California, Berkeley. But in the long term, deep learning may not win the day; some researchers are pursuing other techniques that show promise. “I'm agnostic,” says Malik. “Over time people will decide what works best in different domains.”

Inspired by the brain

Back in the 1950s, when computers were new, the first generation of AI researchers eagerly predicted that fully fledged AI was right around the corner. But that optimism faded as researchers began to grasp the vast complexity of real-world knowledge — particularly when it came to perceptual problems such as what makes a face a human face, rather than a mask or a monkey face. Hundreds of researchers and graduate students spent decades hand-coding rules about all the different features that computers needed to identify objects. “Coming up with features is difficult, time consuming and requires expert knowledge,” says Ng. “You have to ask if there's a better way.”

IMAGES: ANDREW NG

 In the 1980s, one better way seemed to be deep learning in neural networks. These systems promised to learn their own rules from scratch, and offered the pleasing symmetry of using brain-inspired mechanics to achieve brain-like function. The strategy called for simulated neurons to be organized into several layers. Give such a system a picture and

• the first layer of learning will simply notice all the dark and light pixels. 
• The next layer might realize that some of these pixels form edges; 
• the next might distinguish between horizontal and vertical lines. 
• Eventually, a layer might recognize eyes, and might realize that two eyes are usually present in a human face (see 'Facial recognition').

The first deep-learning programs did not perform any better than simpler systems, says Malik. Plus, they were tricky to work with. “Neural nets were always a delicate art to manage. There is some black magic involved,” he says. The networks needed a rich stream of examples to learn from — like a baby gathering information about the world. In the 1980s and 1990s, there was not much digital information available, and it took too long for computers to crunch through what did exist. Applications were rare. One of the few was a technique — developed by LeCun — that is now used by banks to read handwritten cheques.

By the 2000s, however, advocates such as LeCun and his former supervisor, computer scientist Geoffrey Hinton of the University of Toronto in Canada, were convinced that increases in computing power and an explosion of digital data meant that it was time for a renewed push. “We wanted to show the world that these deep neural networks were really useful and could really help,” says George Dahl, a current student of Hinton's.

As a start, Hinton, Dahl and several others tackled the difficult but commercially important task of speech recognition. In 2009, the researchers reported2 that after training on a classic data set — three hours of taped and transcribed speech — their deep-learning neural network had broken the record for accuracy in turning the spoken word into typed text, a record that had not shifted much in a decade with the standard, rules-based approach. The achievement caught the attention of major players in the smartphone market, says Dahl, who took the technique to Microsoft during an internship. “In a couple of years they all switched to deep learning.” For example, the iPhone's voice-activated digital assistant, Siri, relies on deep learning.

Giant leap
When Google adopted deep-learning-based speech recognition in its Android smartphone operating system, it achieved a 25% reduction in word errors. “That's the kind of drop you expect to take ten years to achieve,” says Hinton — a reflection of just how difficult it has been to make progress in this area. “That's like ten breakthroughs all together.”

Meanwhile, Ng had convinced Google to let him use its data and computers on what became Google Brain. The project's ability to spot cats was a compelling (but not, on its own, commercially viable) demonstration of unsupervised learning — the most difficult learning task, because the input comes without any explanatory information such as names, titles or categories. But Ng soon became troubled that few researchers outside Google had the tools to work on deep learning. “After many of my talks,” he says, “depressed graduate students would come up to me and say: 'I don't have 1,000 computers lying around, can I even research this?'”

So back at Stanford, Ng started developing bigger, cheaper deep-learning networks using graphics processing units (GPUs) — the super-fast chips developed for home-computer gaming3. Others were doing the same. “For about US$100,000 in hardware, we can build an 11-billion-connection network, with 64 GPUs,” says Ng.

Victorious machine

But winning over computer-vision scientists would take more: they wanted to see gains on standardized tests. Malik remembers that Hinton asked him: “You're a sceptic. What would convince you?” Malik replied that a victory in the internationally renowned ImageNet competition might do the trick.

In that competition, teams train computer programs on a data set of about 1 million images that have each been manually labelled with a category. After training, the programs are tested by getting them to suggest labels for similar images that they have never seen before. They are given five guesses for each test image; if the right answer is not one of those five, the test counts as an error. Past winners had typically erred about 25% of the time. In 2012, Hinton's lab entered the first ever competitor to use deep learning. It had an error rate of just 15% (ref. 4).

“Deep learning stomped on everything else,” says LeCun, who was not part of that team. The win landed Hinton a part-time job at Google, and the company used the program to update its Google+ photo-search software in May 2013.

Malik was won over. “In science you have to be swayed by empirical evidence, and this was clear evidence,” he says. Since then, he has adapted the technique to beat the record in another visual-recognition competition5. Many others have followed: in 2013, all entrants to the ImageNet competition used deep learning.

“Over the next few years we'll see a feeding frenzy. Lots of people will jump on the deep-learning bandwagon.”

With triumphs in hand for image and speech recognition, there is now increasing interest in applying deep learning to natural-language understanding — comprehending human discourse well enough to rephrase or answer questions, for example — and to translation from one language to another. Again, these are currently done using hand-coded rules and statistical analysis of known text. The state-of-the-art of such techniques can be seen in software such as Google Translate, which can produce results that are comprehensible (if sometimes comical) but nowhere near as good as a smooth human translation. “Deep learning will have a chance to do something much better than the current practice here,” says crowd-sourcing expert Luis von Ahn, whose company Duolingo, based in Pittsburgh, Pennsylvania, relies on humans, not computers, to translate text. “The one thing everyone agrees on is that it's time to try something different.”

Deep science

In the meantime, deep learning has been proving useful for a variety of scientific tasks. “Deep nets are really good at finding patterns in data sets,” says Hinton. In 2012, the pharmaceutical company Merck offered a prize to whoever could beat its best programs for helping to predict useful drug candidates. The task was to trawl through database entries on more than 30,000 small molecules, each of which had thousands of numerical chemical-property descriptors, and to try to predict how each one acted on 15 different target molecules. Dahl and his colleagues won $22,000 with a deep-learning system. “We improved on Merck's baseline by about 15%,” he says.

Biologists and computational researchers including Sebastian Seung of the Massachusetts Institute of Technology in Cambridge are using deep learning to help them to analyse three-dimensional images of brain slices. Such images contain a tangle of lines that represent the connections between neurons; these need to be identified so they can be mapped and counted. In the past, undergraduates have been enlisted to trace out the lines, but automating the process is the only way to deal with the billions of connections that are expected to turn up as such projects continue. Deep learning seems to be the best way to automate. Seung is currently using a deep-learning program to map neurons in a large chunk of the retina, then forwarding the results to be proofread by volunteers in a crowd-sourced online game called EyeWire.

“Deep learning has the property that if you feed it more data, it gets better and better.”

William Stafford Noble, a computer scientist at the University of Washington in Seattle, has used deep learning to teach a program to look at a string of amino acids and predict the structure of the resulting protein — whether various portions will form a helix or a loop, for example, or how easy it will be for a solvent to sneak into gaps in the structure. Noble has so far trained his program on one small data set, and over the coming months he will move on to the Protein Data Bank: a global repository that currently contains nearly 100,000 structures.

For computer scientists, deep learning could earn big profits: Dahl is thinking about start-up opportunities, and LeCun was hired last month to head a new AI department at Facebook. The technique holds the promise of practical success for AI. “Deep learning happens to have the property that if you feed it more data it gets better and better,” notes Ng. “Deep-learning algorithms aren't the only ones like that, but they're arguably the best — certainly the easiest. That's why it has huge promise for the future.”

Not all researchers are so committed to the idea. Oren Etzioni, director of the Allen Institute for Artificial Intelligence in Seattle, which launched last September with the aim of developing AI, says he will not be using the brain for inspiration. “It's like when we invented flight,” he says; the most successful designs for aeroplanes were not modelled on bird biology. Etzioni's specific goal is to invent a computer that, when given a stack of scanned textbooks, can pass standardized elementary-school science tests (ramping up eventually to pre-university exams). To pass the tests, a computer must be able to read and understand diagrams and text. How the Allen Institute will make that happen is undecided as yet — but for Etzioni, neural networks and deep learning are not at the top of the list.

One competing idea is to rely on a computer that can reason on the basis of inputted facts, rather than trying to learn its own facts from scratch. So it might be programmed with assertions such as 'all girls are people'. Then, when it is presented with a text that mentions a girl, the computer could deduce that the girl in question is a person. Thousands, if not millions, of such facts are required to cover even ordinary, common-sense knowledge about the world. But it is roughly what went into IBM's Watson computer, which famously won a match of the television game show Jeopardy against top human competitors in 2011. Even so, IBM's Watson Solutions has an experimental interest in deep learning for improving pattern recognition, says Rob High, chief technology officer for the company, which is based in Austin, Texas.

Google, too, is hedging its bets. Although its latest advances in picture tagging are based on Hinton's deep-learning networks, it has other departments with a wider remit. In December 2012, it hired futurist Ray Kurzweil to pursue various ways for computers to learn from experience — using techniques including but not limited to deep learning. Last May, Google acquired a quantum computer made by D-Wave in Burnaby, Canada (see Nature 498, 286–288; 2013). This computer holds promise for non-AI tasks such as difficult mathematical computations — although it could, theoretically, be applied to deep learning.

Despite its successes, deep learning is still in its infancy. “It's part of the future,” says Dahl. “In a way it's amazing we've done so much with so little.” And, he adds, “we've barely begun”. Nature505,146–148(09 January 2014)doi:10.1038/505146a

References Le, Q. V. et al. Preprint at http://arxiv.org/abs/1112.6209 (2011). Show context
Mohamed, A. et al. 2011 IEEE Int. Conf. Acoustics Speech Signal Process. http://dx.doi.org/10.1109/ICASSP.2011.5947494 (2011). Show context
Coates, A. et al. J. Machine Learn. Res. Workshop Conf. Proc. 28, 1337–1345 (2013). Show context
Krizhevsky, A., Sutskever, I. & Hinton, G. E. In Advances in Neural Information Processing Systems 25; available at http://go.nature.com/ibace6 Show context
Girshick, R., Donahue, J., Darrell, T. & Malik, J. Preprint at http://arxiv.org/abs/1311.2524 (2013).

Deep-learning computers close in on artificial intelligence

Related stories
Neuroelectronics: Smart connections
Computing: The quantum company
Artificial intelligence finds fossil sites

More related stories

ORIGINAL: Nature
Nicola Jones
08 January 2014

Scientists discover double meaning in genetic code

Scientists have discovered a second code hiding within DNA. This second code contains information that changes how scientists read the instructions contained in DNA and interpret mutations to make sense of health and disease.

Genome scientist Dr. John Stamatoyannopolous led
a team that discovered a second code hidden in DNA

A research team led by Dr. John Stamatoyannopoulos, University of Washington associate professor of genome sciences and of medicine, made the discovery. The findings are reported in the Dec. 13 issue of Science. The work is part of the Encyclopedia of DNA Elements Project, also known as ENCODE. The National Human Genome Research Institute funded the multi-year, international effort. ENCODE aims to discover where and how the directions for biological functions are stored in the human genome.

Since the genetic code was deciphered in the 1960s, scientists have assumed that it was used exclusively to write information about proteins. UW scientists were stunned to discover that genomes use the genetic code to write two separate languages.

• One describes how proteins are made, and the 
• other instructs the cell on how genes are controlled. 

One language is written on top of the other, which is why the second language remained hidden for so long.

“For over 40 years we have assumed that DNA changes affecting the genetic code solely impact how proteins are made,” said Stamatoyannopoulos. “Now we know that this basic assumption about reading the human genome missed half of the picture. These new findings highlight that DNA is an incredibly powerful information storage device, which nature has fully exploited in unexpected ways.”

The genetic code uses a 64-letter alphabet called codons. The UW team discovered that some codons, which they called duons, can have two meanings, 

• one related to protein sequence, and 
• one related to gene control. 

These two meanings seem to have evolved in concert with each other. The gene control instructions appear to help stabilize certain beneficial features of proteins and how they are made.
The discovery of duons has major implications for how scientists and physicians interpret a patient’s genome and will open new doors to the diagnosis and treatment of disease.

“The fact that the genetic code can simultaneously write two kinds of information means that many DNA changes that appear to alter protein sequences may actually cause disease by disrupting gene control programs or even both mechanisms simultaneously,” said Stamatoyannopoulos.

Grants from the National Institutes of Health U54HG004592, U54HG007010, and UO1E51156 and National Institute of Diabetes and Digestive and Kidney Diseases FDK095678A funded the research.

In addition to Stamatoyannopoulos, the research team included Andrew B. Stergachis, Eric Haugen, Anthony Shafer, Wenqing Fu, Benjamin Vernot, Alex Reynolds, and Joshua M. Akey, all from the UW Department of Genome Sciences, Anthony Raubitschek of the UW Department of Immunology and Benaroya Research Institute, Steven Ziegler of Benaroya Research Institute, and Emily M. LeProust, formerly of Agilent Technologists and now with Twist Bioscience.

ORIGINAL: U of Washington

UW Health Sciences and UW Medicine

December 12, 2013

Computer simulations reveal the energy landscape of ion channels

ORIGINAL: University of Vienna

Anna Stary-Weinzinger and Tobias Linder, Department of Pharmacology and Toxicology, University of Vienna (Photo private). (Copyright: Anna Stary-Weinzinger and Tobias Linder)

Ion channels are important drug targets. A young team of researchers led by pharmacologist Anna Stary-Weinzinger from the Department of Pharmacology and Toxicology, University of Vienna investigated the opening and closing mechanisms of these channels: For the first time the full energy landscape of such a large protein (> 400 amino acids) can be calculated in atomic detail. The scientists identified a phenylalanine, which plays a key role for the transition between open and closed state. The time consuming calculations were performed using the high performance computer cluster (VSC), which is currently the fastest computer in Austria. Recently, the results were published in PLOS Computational Biology.

Every cell of our body is separated from its environment by a lipid bilayer. In order to maintain their biological function and to transduce signals, special proteins, so called ion channels, are embedded in the membrane. Anna Stary-Weinzinger and Tobias Linder from the University of Vienna and Bert de Groot from the Max Planck Institute of Biophysical Chemistry in Göttingen identified a key amino acid (phenylalanine 114), which plays an essential role for opening and closing of these ion channels. A conformational change of phenylalanine triggers opening of the channels. 

"These proteins are highly selective, they can distinguish between different ions such as sodium, potassium or chloride and allow ion flux rates of up to 100 million ions per seconds", explains Stary-Weinzinger, leader of the research project and postdoc at the Department of Pharmacology and Toxicology of the University of Vienna. "These molecular switches regulate numerous essential body functions such as transduction of nerve signals, regulations of the heart rhythm or release of neurotransmitters. Slight changes in function, caused by replacement of single amino acids, can lead to severe diseases, such as arrhythmias, migraine, diabetes or cancer." 

This is a schematic description of the opening and closing mechanisms of an ion channel. Credit: Anna Stary-Weinzinger 

Knowledge of ion channel function provides the basis for better drugs 

Ion channels are important drug targets. Ten percent of current pharmaceuticals target ion channels. A detailed understanding of these proteins is therefore essential to develop drugs with improved risk-benefit profiles. An important basis for drug development is a detailed knowledge of the functional mechanisms of these channels. However, there are still many open questions; especially the energy profile and pathway of opening and closure are far from being understood. 

Computer simulations visualize ion channel movements 

To watch these fascinating proteins at work, molecular dynamics simulations are necessary. Computational extensive calculations were performed with the help of the Vienna Scientific Cluster (VSC), the fastest high performance computer in Austria, a computer cluster operated by the University of Vienna, the Vienna University of Technology and the University of Natural Resources and Applied Life Sciences Vienna. With the help of VSC, the free energy landscape of ion channel gating could be investigated for the first time. The young researchers discovered that the open and closed channel states are separated by two energy barriers of different height. 

Phenylalanine triggers conformational changes 

Surprisingly, the dynamics of a specific amino acid, phenylalanine 114, are coupled to a first smaller energy barrier. "This side chain acts as molecular switch to release the channel from the closed state," explains Tobias Linder, PhD student from the University of Vienna. After these local changes, the channel undergoes large global rearrangements, leading to a fully open state. This second transition from an intermediate to a fully open pore is accompanied by a large second energy barrier. 

This research project is financed by the FWF-doctoral program "Molecular Drug Targets" (MolTag), which is led by Steffen Hering, Head of the Department of Pharmacology and Toxicology of the Faculty of Life Sciences, University of Vienna. 

PLOS: Probing the energy landscape of activation gating of the bacterial potassium channel KcsA

Source: University of Vienna

Journal reference: PLoS Computational Biology

Provided by University of Vienna

Graphene foams: Cozy and conductive scaffolds for neural stem cells

ORIGINAL: Physorg

by John Hewitt
Apr 04, 2013 

Graphene Foam. Credit: from google, via ars technical 

(Phys.org) —Graphene foams have been around now for a couple years. Their widespread application in everything from electronics and energy storage to substitutes for helium in balloons is still greatly anticipated. Researchers from the Chinese Academy of Sciences in Suzhou, and Beijing, have now shown that graphene foams can also be used to craft conductive scaffolds for neural stem cells. Their open paper, published yesterday in Nature's Scientific Reports, suggests new approaches for neural tissue engineering, and possibly for interfacing with neural prosthetics. 

It has been previously reported that graphene sheets support growth and differentiation of human neural stem cells (NSCs) in a similar fashion to other common substrates like glass or polymer PDMS. Chinese researchers have done pioneering work in synthesizing graphene foams to exacting standards of purity and uniformity. When coated with laminin or other matrix proteins, these foams could potentially serve not only as compatible neural housing but also as a means to control the tenants electrically.

To probe the electrical characteristics of the foam the researchers used cyclic voltammetry, a common technique often used in basic electrochemistry. Their results indicated that the cells could be safely stimulated via capacitive charge injection in the potential window range from -0.2 to + 0.8 V, similar again to results from 2D graphene film studies. They further noted that the 3D foam architecture provides more efficient charge injection and potentially more specific stimulation capability.

It is instructive here to note that our painful collective history with asbestos fiber has shown us that geometry can make the poison as much as any chemical effect. It is not just the aspect of the asbestos fiber, but its inconvenient scale that makes its presence so insidious within the lung. Similarly the researchers did not expect to just toss a few neurons onto a random trusswork and expect the ladder's rungs to be ideally spaced. Indeed the images provided by the authors show the seeded NSCs clinging to the graphene structure like spacewalking astronauts trawling along a space station—but somehow they not only survived, but seemed to thrive.

The graphene foams were synthesized by chemical vapor deposition using a Ni foam template. Scanning electron microscope observation showed a porous structure, which was determined to average 100-300um while the graphene skeleton width was around 100-200um. The surface chemistry of the graphene foams was characterized the X-ray photoelectron spectroscopy (XPS). The criteria used to gauge inertness of the surface was the presence of a large peak corresponding to non-oxygenated rings and small peaks for the C-O bonds. 

NSC adhesion and proliferation on 3D-GF scaffold. Credit: Scientific Reports, doi:10.1038/srep01604 

Cytotoxicity, evaluated using Calcein-AM and EthD-I staining, showed that 90% of the cells were viable at 5 days out. Proliferation of NSCs was measured from the expression of Ki-67 protein, a marker for cell proliferation that is absent during interphase, and was initially expressed in 80% of the cells. After 5 days, the cells exhibited elongated cell shape with neurite outgrowth, and covered the entire foam surface to confluence. Tuj-1-postive neurons, O4-positive oligodendrocytes, and GFAP-positive astrocytes were all observed in healy abundance.

The longer term clinical scenario for these kinds of studies is still unfolding. In the absence of vascularization, neurons can only bear to have so many neighbors nearby, and still receive adequate nourishment through diffusion. The traditional concept of using degradable matrices that would be later implanted into the cortex has yet to be realized. Chemical enticements to integrate with the local neuritic field and vasculature are just beginning to be explored for these kinds of explant studies. Permanent matrices with functionalized surfaces that would also be electrically addressable would be a welcome addition to this toolkit. 

Real cortical gray matter is a jungle where ceaseless competition for every cubic nanometer of space is not just a game of survival, it is the lifeblood of every thought and memory. If you were to imagine wrestlers in a steel cage match, packed to hilt, you would not be too far off. Every electromechanical spike, every mini-potential produced in a dendrite, is a breath. A little extra bit of power upon inhalation to exert upon competitors held in mutual death-grip, only to have it ever so tightened again after each exhalation. Successful introduction of novice and metabolically disadvantaged tissue into this strategic landscape would require certain considerations on its behalf. Extra stimulation, growth factor, or oxygenation might be just what it takes to ensure productive evolution of new structure. 

Another final take-home message of the paper suggests is that some level of editorial patience must have been afforded for the many obvious grammatical missteps and outright phraseological foibles understandably introduced by the Chinese authorship. It is a small price to pay perhaps for our mutual collaboration. It is encouraging that Western journals welcome the continued publication of Chinese advances in fields like graphene processing, along with the efforts of the authors taken to make it understandable to us.

More information: Three-dimensional graphene foam as a biocompatible and conductive scaffold for neural stem cells, Scientific Reports 3, Article number: 1604. doi:10.1038/srep01604

Abstract 

Neural stem cell (NSC) based therapy provides a promising approach for neural regeneration. For the success of NSC clinical application, a scaffold is required to provide three-dimensional (3D) cell growth microenvironments and appropriate synergistic cell guidance cues. Here, we report the first utilization of graphene foam, a 3D porous structure, as a novel scaffold for NSCs in vitro. It was found that three-dimensional graphene foams (3D-GFs) can not only support NSC growth, but also keep cell at an active proliferation state with upregulation of Ki67 expression than that of two-dimensional graphene films. Meanwhile, phenotypic analysis indicated that 3D-GFs can enhance the NSC differentiation towards astrocytes and especially neurons. Furthermore, a good electrical coupling of 3D-GFs with differentiated NSCs for efficient electrical stimulation was observed. Our findings implicate 3D-GFs could offer a powerful platform for NSC research, neural tissue engineering and neural prostheses.

Journal reference: Scientific Reports

Nanotechnology milestone: general method for designing stable proteins

ORIGINAL: Foresight Institute

Comparison of computational models with experimentally determined structure: design model (left) and NMR structure (right). Credit: Nobuyasu Koga et al./Nature)

Yet another milestone along the protein design molecular engineering path to advanced nanotechnology has been reached, thanks to the efforts of the laboratory of David Baker, one of the 2004 winners of the Foresight Feynman Prize in Nanotechnology for Theoretical work. From KurzweilAI “How to design proteins from scratch“:

… By following a set of rules, they designed five proteins from scratch that fold reliably into predicted conformations. In a blind test, the team showed that the synthesized proteins closely match the predicted structures.

“What you have now is a flexible set of building blocks for nanoscale assembly,” says Jeremy England, a molecular biophysicist at the Massachusetts Institute of Technology in Cambridge, who was not involved in the work. …

More detail is given in a commentary in Nature, the journal in which the research was published “Proteins made to order: Researchers design proteins from scratch with predictable structures“:.

… Baker’s proteins are in a sense “platonic ideals”, he says: simple backbone constructs with every amino acid optimized to fold into the prescribed, stable structure. In this way they differ from natural proteins, whose folded structures represent a compromise between the competing requirements of optimum folding and biological function, leading to “frustrated” parts of the sequence that may be essential for function but are destabilizing to the fold. As evidence of their stability, the designed proteins melt at about 100 °C, Koga says, compared to 40–50 °C for a natural protein. …

The best summary of the significance of this result is given by the authors in the conclusion of their paper (reference numbers omitted from quotation):

The design principles and methodology we have described should allow the ready design of a wide range of robust and stable protein building blocks for the next generation of engineered functional proteins. Almost all protein design and engineering efforts so far have repurposed naturally occurring proteins that evolved for some other, often unrelated, function. It should now become possible to custom-design protein scaffolds ideal for the desired function, and to build larger assemblies and materials from robust ideal building blocks.

From the standpoint of advanced nanotechnology/molecular manufacturing, the “build larger assemblies” part is especially interesting. The abstract of the research is available on the journal web site, and the authors have made a full text PDF available on the Baker lab web site.

—James Lewis, PhD

This entry was posted on Wednesday, November 21st, 2012 at 1:13 PM and is filed under Atomically Precise Manufacturing (APM), Bionanotechnology, Computational nanotechnology, Molecular Nanotechnology, Molecular manufacturing, Nano, Nanobiotechnology, Nanotech, Nanotechnology, Productive Nanosystems, Research. 

'Cognitive Big Bang' Discovered in Tiny Sea Worm

ORIGINAL: LiveScience

By Tia Ghose, LiveScience Staff Writer | LiveScience.com

The tiny worm Pikaia gracilens, the earliest known vertebrate ancestor, from the Middle Cambrian of British Columbia, may hold the secret to the expansion of intelligence in vertebrates (animals with backbones). CREDIT: Nobu Tamura, Wikimedia Commons

Gene duplications in the ancient sea worm Pikaia (fossil specimen shown here at the Smithsonian) some 550 million years ago may explain vertebrate smarts.
CREDIT: Public Domain

Several "brainy" genes that were duplicated in a tiny sea creature nearly 550 million years ago may have led to the massive expansion in intelligence in vertebrate species, two new studies have found.

The studies, published today (Dec. 2) in the journal Nature Neuroscience, suggest this duplication of certain genes spurred an explosion in the number of chemicals that regulate brain function in vertebrates (animals with backbones), thereby leading to greater intelligence, the research suggests.

"This genome event produced a kind of cognitive big bang; it produced a large set of interesting behavior," said study co-author Seth Grant, a neuroscientist at the University of Cambridge in the United Kingdom. "It produced a molecular toolbox, which in the case of the brain, produced many, many more proteins that you find in the synapses, the junctions between nerve cells."

The study showed that changes, or mutations, in these genes lead to learning problems in both mice and humans, as well as psychological disorders in humans, said Jeffrey Boore, the CEO of Genome Project Solutions, who was not involved in the study. That supports the notion that these genes "have diversified throughout evolution from their ancient duplications to perform important, specific, diverse roles in mammal cognition in behavior."

Vertebrate explosion

Vertebrates are more intelligent than invertebrates (animals with no backbone), but how those smarts evolved has remained a mystery. In general, vertebrates have many more proteins, which carry out a gene's instructions, in the brain and nervous system; these proteins enable spiny creatures to have a wider, more flexible range of behaviors than invertebrates, Grant said. [Inside the Brain: A Journey Through Time]

"It would be like a stereo or a machine with many more switches in it; it can do more sophisticated types of things," Grant told LiveScience.

Grant and his colleagues wondered whether duplications of certain genes contributed to this explosion in the number of brain proteins, and as a result, to vertebrate intelligence. That's because duplications in genes can give creatures the buffering to evolve more rapidly, Grant said.

(Normally, if a mutation crops up in a gene that's critical to an animal's survival, evolution will weed out that animal, which usually means very little change occurs in critical genes. But if there are two copies of a gene, then the animal has a spare functioning copy, which allows those genes to rack up more mutations — both beneficial, harmful and neutral — without reducing survival fitness.)

In particular, the researchers noticed that vertebrates had multiple copies of two genes that process glutamate, a brain chemical critical for learning and memory. Those genes, which code for the receptor that binds glutamate and the cellular components that process it inside brain cells, were first duplicated in a 2-inch-long (5 centimeters), sea-faring worm called a pikaia that gave rise to all vertebrate species about 550 million yeas ago.

Roots of intelligence

To see whether that duplication formed the genetic origin of intelligence in vertebrates, Grant's team subjected mice with different mutations in these genes to a battery of visual tests that assess learning and attention in changing environments. Different mutations led mice to perform better or worse on cognitive tests, which implied that these genes play a key role in mouse intelligence.

Moreover, the team showed humans and mice carrying the same mutation in a specific glutamate-transporting gene have reduced adaptive learning, suggesting these genes control intelligence and learning across vertebrate species. (In humans, this gene mutation is linked to schizophrenia.)

The findings suggest those first gene duplications probably also gave rise to rapid evolution in the nervous-system proteins that interact with the glutamate receptor, enabling vertebrates to have more subtle, sophisticated responses to their environment.

"Our evidence shows unambiguously that these genome duplications and expansions in the gene family have produced greater complexity to the behavioral repertoire of the vertebrate," Grant said.

Follow LiveScience on Twitter @livescience. We're also on Facebook & Google+. 

New Model Gives Hands-On Help for Learning the Secrets of Molecules

ORIGINAL: ScienceDaily

The models will enable researchers to quickly and collaboratively see, touch, and test ideas about molecular interactions and the behavior of proteins. (Credit: Masaru Kawakami/ Review of Scientific Instruments)

ScienceDaily (Aug. 24, 2012) — For biology researchers, the complex world of molecular proteins -- where tens of thousands of atoms can comprise a single protein -- may be getting clearer with the help of a new soft, transparent, and squishy silicone model they can hold in their hands.

Its advantage over traditional computer and solid models is that it is mostly transparent and easy to manipulate, which will help researchers more intuitively understand protein structures, positions, and interactions. The models will enable researchers to quickly and collaboratively see, touch, and test ideas about molecular interactions and the behavior of proteins. These insights are keys to innovation in drug design because they help generate discussion about what a particular molecular surface might be like and how a protein is shaped and structured.

The models also allow researchers to simulate docking maneuvers involving molecules known as ligands and their partners, a chemical binding step that can turn a biological process on or off.

This boost to molecular modeling comes from Masaru Kawakami, Ph.D., a biophysicist researcher at JAIST (Japan Advanced Institute of Science and Technology) in Ishikawa, Japan. It appears in the current issue of the American Institute of Physics (AIP) journal Review of Scientific Instruments.

"Because my new model is soft, users can deform the model and experience ligand binding or protein-protein association, which has never been possible with other physical molecule models," said Kawakami. "I believe my model would be an effective discussion tool for the classroom or laboratory to stimulate inspired learning."

Building blocks of life came from deep Earth

ORIGINAL: Phys.Org

by Tom Marshall

August 7, 2012 

The rise of the first complex life depended on vital metals brought up to the Earth's surface from far below in vast granite deposits, a new study argues.

Metals like copper, zinc and molybdenum are essential ingredients for certain enzymes and proteins. These are needed for life forms with a complex internal structure, known as eukaryotes, to evolve. Without these metals the history of life could have been very different; plants and animals made of many cells could have taken hundreds of millions more years to develop, if they appeared at all.

The new study's authors realised that eukaryotes started appearing soon after a period of unusual geological activity, and wonder if it could have provided the raw materials they needed.

"Biologists have been saying for a long time that these three metals are essential for complex life to develop," says Professor John Parnell, a geologist at the University of Aberdeen and lead author of the paper, which appears in Geology. "And geologists have been aware that there was a period of unusual geological activity around the same time that would have brought an extraordinary amount of these metals to the surface. But I think we're the first to put the two together and suggest that the geological changes actually enabled the biological advances."

In particular, eukaryotic life is needed for sex differences to emerge. Until living things have both males and females, rapid evolution is impossible; sexual reproduction allows the mixing of genes from both parents, so that a population can contain much more variation for natural selection to work on. Before sexual reproduction, variations in populations of living things could stem only from occasional random mutations, so evolution moved much more slowly. 

The explosion of new life took place during a period known as the Mesoproterozoic, around 1.6 billion years ago. This followed the birth of a new supercontinent known as Nuna or Columbia around 1.9 billion years ago, which triggered major changes in the activity of the Earth's mantle beneath.

Because of the thicker crust below the supercontinent, heat flow at the base of the crust was unusually high, leading to rising magma plumes that brought up metals that had previously been locked deeper in the Earth. After emerging through volcanic activity, this material cooled into vast new fields of granite, with deposits of metal sulphides disseminated throughout. As weathering slowly uncovered these, they turned to sulphates and were washed into rivers, lakes and shallow coastal waters, where they became available for use by living things. With these nutrients in place, the stage was set for the appearance of eukaryotes.

"Metals do come to the surface through normal volcanic processes, but we think this episode of high heat flow greatly accelerated the process. So it's possible that eukaryotes would still have developed if this hadn't happened, but it might have taken a lot longer," says Parnell.

Until recently, scientists thought these metals came instead from changes in the chemistry of the oceans. The discovery is part of a wider move towards theories that complex life got its start on land or in shallow waters rather than the deep ocean, as previously believed.

"I suspect that this increases the focus on the terrestrial origins of eukaryotes, as opposed to the deep marine ones," Parnell comments. "The onus is now on the palaeobiologists to go out and see what traces of early life they can find in the terrestrial record."

Journal reference: Geology

Provided by PlanetEarth Online

This story is republished courtesy of Planet Earth online, a free, companion website to the award-winning magazine Planet Earth published and funded by the Natural Environment Research Council (NERC).

Plants to express human proteins

ORIGINAL: PhyOrg

July 30, 2012  

Credit: Thinkstock

Plants are emerging as important biotechnology tools for the production of highly purified recombinant proteins. The Plastomics network dissected the process of foreign gene insertion into plants and their potential use as protein production reservoirs. 

Transformation of the gene of interest in the plastid genome offers certain advantages compared to introducing it into the nucleus. More than 70 % of the leaf protein will consist of the recombinant protein and transfer of plastid DNA occurs through maternal transmission and not through pollen, providing a strong biological containment. 

Furthermore, gene integration occurs via homologous recombination at specific and predictable sites of the genome with hardly any silencing being reported. The aim of the European Plastomics project was to understand the genes and proteins implicated during plastid transformation of tobacco, tomato and potato. 

The ultimate goal was to improve plastid transformation frequencies, the regulated expression of transgenes in different plastid types and the stability of expressed proteins. Project teams succeeded in identifying the proteins involved in transgene integration into and excision from the plastid genome. The process of gene transcription was further delineated and control elements were generated that improved transcription of foreign genes in tobacco chloroplasts. An important achievement was the development of a cleavable protein-fusion system capable of producing large amounts of interferon 2b in tobacco chloroplasts. 

The information generated during the Plastomics project enhanced our understanding of the processes underlying the insertion and removal of foreign genes into the plastid genome of plants. Certain findings proved to be commercially exploitable and culminated with the application of a patent for a translation control element.

Overall, project results are expected to have significant implications in the rapidly expanding field of plant biotechnology. 

Provided by CORDIS 

Computational Approaches to Elucidating Transient Protein-Protein Interactions, Predicting Receptor-Ligand Pairings

ORIGINAL: IntechOpen

Computational Approaches to Elucidating Transient Protein-Protein Interactions, Predicting Receptor-Ligand Pairings

By Ernesto Iacucci, Samuel Xavier De Souza and Yves Moreau

1. Introduction   

Protein-protein interactions (PPI) are one of the most important biological events which occur in the cell. As PPIs regulate almost all biological processes in the cell, aberrations in PPI may cause severe health problems. One specific area of PPI is receptor-ligand interactions. These interactions are transient yet account for a large part of cell-to-cell communication. As PPI is an important area of research, many groups have proposed methods to make computational predictions of PPI.  The basis of the majority of these methods rely largely on the phylogenetic profile analysis of candidate interactors. These methods determine the similarity of the phylogenetic history of a protein A and its putative protein partner B, examining the most accurate measure of similarity between the phylogenetic histories of A and B in order to predict interaction. As interacting proteins should co-adapt as they are under the same evolutionary pressures, it is self-evident that interacting receptors and ligands should be identifiable by application of the same methodology.   While several methods, described below, make use of phylogenetic information to predict protein-protein interaction (PPI), more contemporary work has been conducted in the area of data fusion and kernel learning. We describe one method [Iacucci et al. 2011] in detail which does both. In this work, the existing line of phylogenetic research is extended by using phylogenetic data to construct a kernel to train a least square support vector machines (LS-SVM) in order to classify candidate receptors and ligands as interacting or noninteracting.    

In this chapter, we discuss the plethora of various methods for determining protein-protein interactions. In addition, we evaluate the application of LS-SVMs to the sub-problem of receptor-ligand interaction prediction.

Fig. 1. The Receptor Ligand Schematic. Schematic of receptor-ligand and protein-protein interaction model. Top image is a representation of in-vivo interaction of proteins, receptors, and ligands while bottom image is the graph representation from which a PPI adjacency matrix may be derived. (Figure published in Iacucci et al. 2010)   

Fig. 2. Phylogenetic Analysis of Proteins

Fig. 3. Work flow of the combined kernel classifier

An expanded genetic alphabet could lead to more easily designed proteins

ORIGINAL: Foresight Institute

Floyd E. Romesberg, associate professor at Scripps Research Institute (Credit: The Scripps Research Institute)

Back in 1992, in chapter 15 of Nanosystems, Eric Drexler suggested that it would be easier to design proteins that fold predictably, an important step on the road to advanced nanotechnology (or molecular manufacturing, or atomically precise manufacturing) if additional amino acids beyond the 20 that are genetically coded could be incorporated into proteins. Chemical synthesis of peptides has provided a way to accomplish this for small amounts of short proteins, but to obtain large amounts of long proteins, it would be very convenient to expand the genetic alphabet to encode additional amino acids. This long-standing effort has taken a major step forward with the discovery of how artificial DNA base pairs can be replicated. A hat tip to ScienceDaily for reprinting this Scripps Research Institute news release “Scripps Research Institute study suggests expanding the genetic alphabet may be easier than previously thought“:

A new study led by scientists at The Scripps Research Institute suggests that the replication process for DNA—the genetic instructions for living organisms that is composed of four bases (C, G, A and T)—is more open to unnatural letters than had previously been thought. An expanded “DNA alphabet” could carry more information than natural DNA, potentially coding for a much wider range of molecules and enabling a variety of powerful applications, from precise molecular probes and nanomachines to useful new life forms.

The new study, which appears in the June 3, 2012 issue of Nature Chemical Biology [abstract], solves the mystery of how a previously identified pair of artificial DNA bases can go through the DNA replication process almost as efficiently as the four natural bases.

“We now know that the efficient replication of our unnatural base pair isn’t a fluke, and also that the replication process is more flexible than had been assumed,” said Floyd E. Romesberg, associate professor at Scripps Research, principal developer of the new DNA bases, and a senior author of the new study. The Romesberg laboratory collaborated on the new study with the laboratory of co-senior author Andreas Marx at the University of Konstanz in Germany, and the laboratory of Tammy J. Dwyer at the University of San Diego.

Adding to the DNA Alphabet

Romesberg and his lab have been trying to find a way to extend the DNA alphabet since the late 1990s. In 2008, they developed the efficiently replicating bases NaM and 5SICS, which come together as a complementary base pair within the DNA helix, much as, in normal DNA, the base adenine (A) pairs with thymine (T), and cytosine (C) pairs with guanine (G).

The following year, Romesberg and colleagues showed that NaM and 5SICS could be efficiently transcribed into RNA in the lab dish. But these bases’ success in mimicking the functionality of natural bases was a bit mysterious. They had been found simply by screening thousands of synthetic nucleotide-like molecules for the ones that were replicated most efficiently. And it had been clear immediately that their chemical structures lack the ability to form the hydrogen bonds that join natural base pairs in DNA. Such bonds had been thought to be an absolute requirement for successful DNA replication-—a process in which a large enzyme, DNA polymerase, moves along a single, unwrapped DNA strand and stitches together the opposing strand, one complementary base at a time.

An early structural study of a very similar base pair in double-helix DNA added to Romesberg’s concerns. The data strongly suggested that NaM and 5SICS do not even approximate the edge-to-edge geometry of natural base pairs—termed the Watson-Crick geometry, after the co-discoverers of the DNA double-helix. Instead, they join in a looser, overlapping, “intercalated” fashion. “Their pairing resembles a ‘mispair,’ such as two identical bases together, which normally wouldn’t be recognized as a valid base pair by the DNA polymerase,” said Denis Malyshev, a graduate student in Romesberg’s lab who was lead author along with Karin Betz of Marx’s lab.

Yet in test after test, the NaM-5SICS pair was efficiently replicable. “We wondered whether we were somehow tricking the DNA polymerase into recognizing it,” said Romesberg. “I didn’t want to pursue the development of applications until we had a clearer picture of what was going on during replication.”

Edge to Edge

To get that clearer picture, Romesberg and his lab turned to Dwyer’s and Marx’s laboratories, which have expertise in finding the atomic structures of DNA in complex with DNA polymerase. Their structural data showed plainly that the NaM-5SICS pair maintain an abnormal, intercalated structure within double-helix DNA—but remarkably adopt the normal, edge-to-edge, “Watson-Crick” positioning when gripped by the polymerase during the crucial moments of DNA replication.

“The DNA polymerase apparently induces this unnatural base pair to form a structure that’s virtually indistinguishable from that of a natural base pair,” said Malyshev.

NaM and 5SICS, lacking hydrogen bonds, are held together in the DNA double-helix by “hydrophobic” forces, which cause certain molecular structures (like those found in oil) to be repelled by water molecules, and thus to cling together in a watery medium. “It’s very possible that these hydrophobic forces have characteristics that enable the flexibility and thus the replicability of the NaM-5SICS base pair,” said Romesberg. “Certainly if their aberrant structure in the double helix were held together by more rigid covalent bonds, they wouldn’t have been able to pop into the correct structure during DNA replication.”

An Arbitrary Choice?

The finding suggests that NaM-5SICS and potentially other, hydrophobically bound base pairs could some day be used to extend the DNA alphabet. It also hints that Evolution’s choice of the existing four-letter DNA alphabet—on this planet—may have been somewhat arbitrary. “It seems that life could have been based on many other genetic systems,” said Romesberg.

He and his laboratory colleagues are now trying to optimize the basic functionality of NaM and 5SICS, and to show that these new bases can work alongside natural bases in the DNA of a living cell.

“If we can get this new base pair to replicate with high efficiency and fidelity in vivo, we’ll have a semi-synthetic organism,” Romesberg said. “The things that one could do with that are pretty mind blowing.”

Now that these scientists have demonstrated that DNA replication is far more flexible than had been thought, it will be fascinating to see what researchers do to expand the genetic alphabet, what additional amino acids they choose to incorporate into what proteins, and what this expansion of the set of amino acids composing proteins means for predicable protein folding and for artificial protein molecular machines.

—James Lewis, PhD