UCT
Austin, TX, United States
Austin, TX, United States

The University of Cape Town is a public research university located in Cape Town in the Western Cape province of South Africa. UCT was founded in 1829 as the South African College, and is the oldest university in South Africa and the second oldest extant university in Africa. The language of instruction is English. Wikipedia.


Time filter

Source Type

News Article | December 6, 2016
Site: www.eurekalert.org

BMJ, one of the world's leading healthcare knowledge providers, and its partner the University of Cape Town Lung Institute's Knowledge Translation Unit, have launched the global edition of the Practical Approach to Care Kit (PACK) in eBook and print format - to support and empower primary healthcare workers. The PACK Adult Global guide provides a generic 'framework' that can be customised to meet the needs of primary healthcare systems in individual countries or states. Primary healthcare is key to achieving the United Nations-led Sustainable Development Goals and the broader goal of "health for all" by providing accessible, affordable and effective healthcare. Yet in many low and middle income countries, primary healthcare is constrained by a lack of adequately skilled and supervised health workers. The Knowledge Translation Unit is a health systems research unit that has spent 16 years developing the PACK programme to empower health workers in primary healthcare. The PACK programme consists of 4 pillars: 1) the PACK Guide 2) the PACK training programme, 3) health systems intervention and strengthening and 4) monitoring and evaluation of the PACK implementation. The PACK Guide is a clinical decision tool that enables healthcare practitioners to manage symptoms and diagnose conditions commonly seen in primary care including infectious diseases, non-communicable diseases, women's health, mental health and end-of-life care. Its comprehensive scope promotes delivery of integrated primary care, rather than care through "vertical programmes". PACK is based on WHO guidelines, strengthened with the latest global research evidence as appraised, graded, sourced and synthesised by BMJ Best Practice. BMJ Best Practice content underpins more than 80% of PACK recommendations and its robust updating processes will facilitate an annual update of the PACK Guide. The PACK programme has been implemented and scaled up throughout the nine provinces in South Africa. It is now used in more than 2,000 clinics across the country, by over 20,000 primary care health workers. The Knowledge Translation Unit has conducted formal evaluation of the programme through 4 pragmatic trials and has published the outcomes of this research in 11 papers. There has been considerable interest from countries outside of South Africa in adopting the PACK programme to strengthen primary care service delivery. PACK is designed to be localised to local clinical protocols, policy and practice, and where necessary translated. PACK has previously been localised for Botswana and Malawi and currently pilot implementations are underway in Florianopolis, Brazil and in three states of Nigeria. The Knowledge Translation Unit has developed a mentorship package to support the in-country localisation of PACK. BMJ and the Knowledge Translation Unit partnered in 2015 to address this need, and to enable global use of the PACK programme in keeping with their strongly aligned strategic goals of "improving primary care, especially in underserved communities" and "enabling a Healthier World". It is to this end that we launch PACK Global Adult 2016/17. Dr Tracy Eastman, Director of PACK Global Development at BMJ said: "The growing burden of non-communicable disease and mental illness in low and middle income countries means it is crucial that healthcare workers have access to the latest evidence based recommendations to diagnose and treat patients effectively across the full spectrum of integrated primary care needs and services. We are delighted to be able to provide the global edition of the PACK Adult guide that can be customised to help meet the specific needs of local communities, as part of the BMJ's Global Health outreach programme." Professor Lara Fairall, Head of the KTU, commented: "The KTU team is committed to improving primary healthcare, focusing on the most underserved communities globally. We hope that by launching the PACK Adult Global guide as an eBook this will raise awareness of the PACK programme and how it can empower primary care clinicians. We encourage primary healthcare workers to utilise the PACK principles and approach to support their teams, strengthen their health systems and provide the highest quality, integrated, team based primary healthcare services, no matter where they work." The UCT Lung Institute Knowledge Translation unit is generously supported by the Peter Sowerby Charitable Foundation. BMJ is a healthcare knowledge provider that aims to advance healthcare worldwide by sharing knowledge and expertise to improve experiences, outcomes and value. For a full list of BMJ products and services, please visit: bmj.com. Follow us on Twitter The KTU is a research unit committed to improving the quality of primary healthcare for underserved communities through pragmatic research, evidence-based implementation, evaluation, and engagement of health systems, their planners and providers. Further details are provided at http://www. The Peter Sowerby Foundation was established in 2011 with an endowment from Dr Peter Sowerby, a retired Yorkshire GP and co-founder of Egton Medical Information Systems, which provides database software to around half of the GP practices in the country. The Foundation does not solicit applications, but seeks out projects to support on ways to improve innovation in primary healthcare, as well as work promoting environmental conservation and activities in Peter's native North Yorkshire. For more information see http://www.


News Article | November 16, 2016
Site: phys.org

The Vela supercluster in its wider surroundings: The image displays the smoothed redshift distribution of galaxies in and around the Vela supercluster (larger ellipse; VSC). The centre of the image, so-called the Zone of Avoidance, is covered by the Milky Way (with its stellar fields and dust layers shown in grey scale), which obscures all structures behind it. Colour indicates the distance ranges of all galaxies within 500 - 1000 million light years (yellow is close to the peak of the Vela supercluster, green is nearer and orange further away). The ellipse marks the approximate extent of the Vela Supercluster, crossing the Galactic Plane. The VSC structure was revealed thanks to the new low latitude spectroscopic redshifts. Given its prominence on either side of the plane of the Milky Way it would be highly unlikely for these cosmic large-scale structures not to be connected across the Galactic Plane. The structure may be similar in aggregate mass to the Shapley Concentration (SC, smaller ellipse), although much more extended. The so-called “Great Attractor” (GA), located much closer to the Milky Way, is an example of a large web structure that crosses the Galactic Plane, although much smaller in extent than VSC. The central, dust-shrouded part of the VSC remains unmapped in the current Vela survey. Also visible are the Milky Way’s two satellite galaxies, LMC and SMC, located south of the Galactic plane. Credit: Thomas Jarrett (UCT) An international team of astronomers has discovered a previously unknown major concentration of galaxies in the constellation Vela, which they have dubbed the Vela supercluster. The gravitational attraction from this large mass concentration in our cosmic neighbourhood may have an important effect on the motion of our Local Group of galaxies including the Milky Way. It may also help to explain the direction and amplitude of the Local Group's peculiar velocity with respect to the Cosmic Microwave Background. Superclusters are the largest and most massive known structures in the Universe. They consist of clusters of galaxies and walls that span up to 200 million light-years across the sky. The most famous supercluster is the Shapley Supercluster, some 650 million light-years away containing two dozens of massive X-ray clusters for which thousands of galaxy velocities have been measured. It is believed to be the largest of its kind in our cosmic neighbourhood. Now a team from South Africa, the Netherlands, Germany, and Australia including two scientists at the Max-Planck-Institut für extraterrestrische Physik in Garching, has discovered another major supercluster, slightly further away (800 million light-years distant), which covers an even larger sky area than Shapley. The Vela supercluster had gone unnoticed due to its location behind the plane of the Milky Way, where dust and stars obscure background galaxies, resulting in a broad band void of extragalactic sources. The team's results suggest the Vela supercluster might be as massive as Shapley, which indicates that its influence on local bulk flows is comparable to that of Shapley. The discovery was based on multi-object spectroscopic observations of thousands of partly obscured galaxies. Observations in 2012 with the refurbished spectrograph of the Southern African Large Telescope (SALT) confirmed that eight new clusters reside within the Vela area. Subsequent spectroscopic observations with the Anglo-Australian Telescope in Australia provided thousands of galaxy redshifts and revealed the vast extent of this new structure. Prof Renée Kraan-Korteweg from the University of Cape Town, who led this study and has been investigating this region for more than a decade, says: "I could not believe such a major structure would pop up so prominently," when she and her colleagues analysed the spectra of the new survey. Scientists Hans Böhringer and Gayoung Chon from the Max-Planck-Institut für extraterrestrische Physik in Garching have surveyed the supercluster region for X-ray luminous galaxy clusters and found two massive clusters in the region covered by the redshift survey and further massive clusters in the immediate vicinity. They thus confirm: "This discovery shows that the Vela supercluster has a significantly higher matter density than average, making it a prominent large structure." But there is still much to do – further follow-up observations are needed to unveil the full extent, mass, and influence of the Vela supercluster. So far this region of the sky is sparsely sampled, while the part closest to the Milky Way has not been probed because dense star and dust layers block our view. Planned observations with the new radio astronomical facility MeerKAT will in particular help to map this obscured region and further optical redshifts will be obtained with the new large-field-of-view multiobject-spectrograph, Taipan, from Australia. The ongoing survey of X-ray luminous clusters conducted by the MPE team, Hans Böhringer and Gayoung Chon, has recently been extended to cover this region in the band of the Milky Way. The area of the Vela supercluster and its environment will receive special attention. "We already have good indications that the Vela Supercluster is embedded in a large network of cosmic filaments traced by clusters, providing insight into the even larger-scale structure embedding the Vela Supercluster. With the future multi-wavelength programme we hope to unveil its full influence on the cosmography and cosmology," Gayoung Chon remarks. Explore further: Laniakea: Newly identified galactic supercluster is home to the Milky Way More information: Renée C. Kraan-Korteweg et al. Discovery of a supercluster in the ZOA in Vela, Monthly Notices of the Royal Astronomical Society: Letters (2016). DOI: 10.1093/mnrasl/slw229


HOUSTON--(BUSINESS WIRE)--The 2017 Underground Construction Technology International Conference & Exhibition (UCT) is taking its benchmark education program to the Fort Worth Convention Center, Jan. 31-Feb. 2. Presented by Underground Construction magazine, UCT is the largest event in the United States focusing on underground utility infrastructure rehabilitation and construction. Conference attendees will learn the latest in underground utility pipe rehabilitation and new construction using both trenchless and open-cut technologies through hands-on demonstrations in the exhibit hall or by real-world case histories, presentations and panel discussions in the seminars. The educational program offers 27 Professional Development Hours, reviewed and certified by The University of Texas at Arlington (Continuing Education Units are also available). Attendee demographics include public works, telecom, gas and electric, government officials, contractors, engineers, manufacturers and suppliers. A wide variety of technologies are highlighted at the conference such as horizontal directional drilling, pipe bursting, cured-in-place pipe and dozens more. The city of Fort Worth’s Water Director, John Robert Carman, will deliver the welcome and keynote address at 9 a.m. on Tuesday, Jan. 31, immediately following a continental breakfast for municipal personnel, contractors, engineers and vendors. Academic sponsors lending their expertise to the program include: the Center for Underground Infrastructure Research and Education, University of Texas at Arlington; the Trenchless Technology Center, Louisiana Tech University; the Center for Innovative Grouting Materials & Technology, University of Houston; Vanderbilt University; Colorado School of Mines; Swim Center at Virginia Tech University; Del E. Webb School of Construction, Arizona State University; the Centre for Advancement of Trenchless Technologies, University of Waterloo (Canada); and Oklahoma State University. UCT also has the support of industry associations such as the National Association of Sewer Service Companies, Distribution Contractors Association, Power & Communication Contractors Association, NACE International, Pipe Line Contractors Association, Interstate Natural Gas Association of America, North American Society For Trenchless Technology, American Gas Association, Southern Gas Association and many more. A limited number of exhibit booths and sponsor opportunities remain. Press registration is complimentary. Early registration discounts are available for multiple attendees from the same company. More information is available at uctonline.com, or contact Karen Francis at kfrancis@uctonline.com. UCT and Underground Construction are produced and managed by Oildom Publishing Company of Texas. Oildom Publishing produces directories, events, magazines and webinars, focused on the energy pipeline and underground utility industry.


A mysterious alignment has been witnessed in a remote area of the universe. Sixty-four supermassive black holes have been observed to be spinning out radio jets from their centers, all pointing towards the same direction. Black holes are well known to emit radio emissions. However, this is the first time their alignment is of such a great magnitude. This phenomenon implies that the force governing these black holes is much greater and older, hence the alignment has been linked to "primordial mass fluctuations" in the early universe. "Since these black holes don't know about each other, or have any way of exchanging information or influencing each other directly over such vast scales, this spin alignment must have occurred during the formation of the galaxies in the early universe," said Professor Andrew Russ Taylor, joint UWC/UCT SKA Chair, Director of the recently launched Inter-University Institute for Data Intensive Astronomy, and principal author of the Monthly Notices study. The astronomers have been puzzled over this alignment and have speculated a few theories that could have been responsible for triggering this large scale phenomenon. Few of the speculated theories include cosmic strings – theoretical fault lines in the universe, exotic particles like axions or cosmic magnetic fields, or maybe something entirely different altogether, which is yet to be ascertained. Experts said the recent observation of black hole alignment could provide evidence of the environmental influences that contributed to the formation and evolution of galaxies as well as the primordial fluctuations that brought about the structure of the universe. This strange phenomenon was captured as a result of three years of deep radio imaging carried out by the Giant Metrewave Radio Telescope (GMRT) located in India. The alignment may hold clues about the early universe when the black holes had initially formed. The study was published in the Monthly Notices of the Royal Astronomical Society. © 2016 Tech Times, All rights reserved. Do not reproduce without permission.


News Article | March 29, 2016
Site: www.biosciencetechnology.com

An international team of scientists, including groups from UC San Francisco, Gladstone Institutes, and the University of Cape Town (UCT), South Africa, have for the first time identified genes and gene regulatory elements that are essential in wing development in the Natal long-fingered bat (Miniopterus natalensis), a species widely distributed in eastern and southern Africa. The new research — presented in two papers published on March 28, one in Nature Genetics and one in PLoS Genetics — revealed regulatory switches that turn bat genes on and off at crucial times during limb development, and has implications for understanding how differences in the size, shape and structure of limbs are generated in mammals in general, including humans, the researchers said. “This gives us our first detailed picture of the genomics behind bat wing development,” said co-senior investigator Nadav Ahituv, Ph.D., a UCSF associate professor of bioengineering and therapeutic sciences in the UCSF School of Pharmacy and member of the UCSF Institute for Human Genetics, whose lab also studies the genetics of human limb malformations. “While some attempts have been made to identify the molecular events that led to the evolution of the bat wing, these have been primarily done on a ‘gene by gene’ basis. In contrast, this work lays out a genome-wide blueprint for the causes that led to the development of the bat wing, a key evolutionary innovation that contributed to bats becoming the second most diverse order of mammals.” Bats are the only mammals capable of powered flight — an innovation that is thought to have occurred about 50 million years ago. Biologists since Charles Darwin have used the structure of the bat wing as an example of both evolutionary novelty — the appearance of a new trait — and vertebrate homology, or shared ancestry between two seemingly different structures — in this case, the wing of the bat and the forelimb of other mammals. But the path of bats’ unique evolution is unclear, noted Nicola Illing, Ph.D., co-senior investigator in the Department of Molecular and Cell Biology at UCT. “The fossil record does not show the transition from tree-climbing mammals with short, free digits to ones that have elongated fingers supporting a wing,” Illing said. “Until now, scientists knew very little about how genes are turned on and off during bat embryonic development to transform a mammalian forelimb into a wing.” In the Nature Genetics paper, the scientists, including co-lead authors Walter L. Eckalbar, Ph.D., a postdoctoral fellow in Ahituv’s laboratory at UCSF, and Ph.D. student Stephen Schlebusch of UCT, first sequenced the entire genome of the Natal long-fingered bat. They then performed detailed molecular genomic analysis on bat embryos collected by Illing and her research group at the de Hoop Nature Reserve in South Africa. The researchers identified over 7,000 genes that are expressed differently in forelimbs compared with hindlimbs at three key stages of bat wing development. They found that many signaling pathways are activated differentially as well, including pathways important in limb formation, digit growth, long bone development and cell death. Also expressed differently are many proteins associated with ribosomes – molecular machines found in all cells that are responsible for protein production during limb development. “It took bats millions of years to evolve wings,” said Eckalbar. “Our work shows that they did this through thousands of genetic alterations, involving both genes used by all animals during limb development and genes whose usage in limb development may be unique to bats.” In addition, the scientists found thousands of genetic switches, called enhancers, which regulate the timing and levels of gene expression and show differences in activity between forelimbs and hindlimbs at these key stages of wing development. “Importantly, this work identified not just which genes are expressed at what stage of growth, but the genetic switches in the genome that are responsible for turning those genes on and off,” Ahituv said. In the study published in PLoS Genetics, the research team, including co-lead authors Betty M. Booker, Ph.D., a post-doctoral fellow in Ahituv’s laboratory, and Tara Friedrich, a Ph.D. student at UCSF and Gladstone Institutes, searched for the evolutionary origin of the bat wing. “We identified genomic sequences that have not changed in most vertebrates, but experienced rapid changes in the common ancestor of today’s bats,” explained Friedrich, a member of the laboratory of co-senior investigator Katherine S. Pollard, Ph.D., a senior investigator at the Gladstone Institutes, a UCSF professor of epidemiology and biostatistics, and a member of the UCSF Institute for Human Genetics. The team mapped these so-called “bat accelerated regions” (BARs) onto areas that were predicted to be important switches that turn genes on during limb development, and found 166 BARs with the potential to influence bat wing development. The researchers tested the effects of five of these BARs in genetically modified mouse embryos and found that all five bat sequences were capable of switching on a reporter gene in the developing mouse forelimb. They noted that one region, BAR116, is located near the HoxD genes, which are known to be involved in limb patterning and skeletal growth. Previously, Mandy Mason, a Ph.D. student at UCT, had shown that two of the HoxD genes — Hoxd10 and Hoxd11 — are far more active in bat wings compared to bat legs during their embryonic development. Following up these lines of evidence, the researchers showed that the bat BAR116 sequence appears to function as a genetic switch that is active in developing limbs, in particular the forelimbs, while the equivalent mouse sequence did not show any activity. “Our computational method enabled identification of DNA sequences that changed dramatically during the emergence of bats,” said Pollard. “It is exciting to see that this evolutionary signature pointed us to parts of the mammalian genome that control limb development.” In addition to unveiling new fundamental details of the evolutionary and developmental origins of powered flight in bats, the new research may provide broader insights into the biological processes that control how mammalian limbs develop in general, Ahituv said. “Importantly, this work will increase our understanding of how alterations in limb development could lead to limb malformations in humans,” he said. “Potentially, it could eventually help contribute to the development of tools and techniques to prevent such malformations.”


News Article | October 26, 2016
Site: www.nature.com

A heavy security presence awaits academics at the University of KwaZulu-Natal campus in Westville, South Africa. After a library at the university’s nearby Durban campus was torched last month, police officers now regularly search staff and their cars for petrol bombs, says Kavilan Moodley, an astrophysicist there. Still, the institution remains open for research and teaching. On the other side of the country, all classes and lab-based research at the Cape Peninsula University of Technology (CPUT) in Cape Town have ground to a halt. The lockdown follows a non-fatal attack late on 11 October in which three men were locked in a university building that was then set on fire. “The physical lockdown has been about a week. Since the arson attacks, we could not guarantee staff and students’ safety,” says Mellet Moll, assistant dean for research in the engineering faculty. Campus violence is affecting many of South Africa’s 26 universities — and the impact is spilling over into research. Student protests against rising tuition fees began in 2015 as the #FeesMustFall campaign, which secured a freeze on fees for this academic year. But the protests flared up again and have become increasingly physically destructive since September, when the government announced that an 8% hike in fees would be permitted for the 2017 academic year. Many undergraduate classes around the country have been stopped, including those at the University of Cape Town (UCT), where face-to-face teaching has been suspended in all faculties. Last week, two security guards were attacked. The effects on research are uneven. Scientists at the UCT, the University of the Witwatersrand in Johannesburg and the University of Pretoria say that, despite the distraction of protests and security, and difficulties in getting packages delivered, they are able to continue with their work. At the CPUT and the University of the Western Cape (UWC) in Cape Town, however, protests have been catastrophic for research. A senior academic at the UWC says that the situation is dire. With the university closed, funding is going unspent, causing international study visas and bursaries for master’s and doctoral students to expire with research not yet complete. University administrators are in a bind. Steadily rising fees have enraged students, who connect the issue to the social and racial disadvantage that persists two decades after the end of apartheid. But higher-education institutions say that they rely on those fees to make up for declining government subsidies. “It’s a crisis we are all facing,” Moll says. “But it’s something that the universities themselves cannot do a lot about. We’ve become the battle-ground between the government and the youth.” Academics are also worried that, in the longer term, the violence could damage universities’ reputations — which could put off foreign students and international collaborators, even at institutions that have not otherwise been affected by the protests. “I’m concerned that if we don’t resolve this within a reasonable timescale, we’ll be seen as dysfunctional, even if it’s not true,” says Don Cowan, director of the University of Pretoria’s Genomics Research Institute. Researchers are also concerned at the prospect of falling budgets as the government — struggling with a combination of the student crisis, the country’s worst drought in decades and slow economic growth — looks to trim other areas of spending. The South African Medical Research Council, for example, has been given a 7% budget cut for the year 2017–18, says council head Glenda Gray. For now, undergraduates are the main concern: universities across the country are holding emergency meetings, general assemblies and peace accords in a desperate bid to keep the academic year alive. “We are now reaching the ‘point of no return’ in terms of saving the academic year,” said UCT vice-chancellor Max Price in a notice to students and staff last week.


News Article | March 29, 2016
Site: phys.org

The new research—presented in two papers published on March 28, one in Nature Genetics and one in PLoS Genetics—revealed regulatory switches that turn bat genes on and off at crucial times during limb development, and has implications for understanding how differences in the size, shape and structure of limbs are generated in mammals in general, including humans, the researchers said. "This gives us our first detailed picture of the genomics behind bat wing development," said co-senior investigator Nadav Ahituv, PhD, a UCSF associate professor of bioengineering and therapeutic sciences in the UCSF School of Pharmacy and member of the UCSF Institute for Human Genetics, whose lab also studies the genetics of human limb malformations. "While some attempts have been made to identify the molecular events that led to the evolution of the bat wing, these have been primarily done on a 'gene by gene' basis. In contrast, this work lays out a genome-wide blueprint for the causes that led to the development of the bat wing, a key evolutionary innovation that contributed to bats becoming the second most diverse order of mammals." Bats are the only mammals capable of powered flight—an innovation that is thought to have occurred about 50 million years ago. Biologists since Charles Darwin have used the structure of the bat wing as an example of both evolutionary novelty—the appearance of a new trait—and vertebrate homology, or shared ancestry between two seemingly different structures—in this case, the wing of the bat and the forelimb of other mammals. But the path of bats' unique evolution is unclear, noted Nicola Illing, PhD, co-senior investigator in the Department of Molecular and Cell Biology at UCT. "The fossil record does not show the transition from tree-climbing mammals with short, free digits to ones that have elongated fingers supporting a wing," Illing said. "Until now, scientists knew very little about how genes are turned on and off during bat embryonic development to transform a mammalian forelimb into a wing." In the Nature Genetics paper, the scientists, including co-lead authors Walter L. Eckalbar, PhD, a postdoctoral fellow in Ahituv's laboratory at UCSF, and PhD student Stephen Schlebusch of UCT, first sequenced the entire genome of the Natal long-fingered bat. They then performed detailed molecular genomic analysis on bat embryos collected by Illing and her research group at the de Hoop Nature Reserve in South Africa. The researchers identified over 7,000 genes that are expressed differently in forelimbs compared with hindlimbs at three key stages of bat wing development. They found that many signaling pathways are activated differentially as well, including pathways important in limb formation, digit growth, long bone development and cell death. Also expressed differently are many proteins associated with ribosomes – molecular machines found in all cells that are responsible for protein production during limb development. "It took bats millions of years to evolve wings," said Eckalbar. "Our work shows that they did this through thousands of genetic alterations, involving both genes used by all animals during limb development and genes whose usage in limb development may be unique to bats." In addition, the scientists found thousands of genetic switches, called enhancers, which regulate the timing and levels of gene expression and show differences in activity between forelimbs and hindlimbs at these key stages of wing development. "Importantly, this work identified not just which genes are expressed at what stage of growth, but the genetic switches in the genome that are responsible for turning those genes on and off," Ahituv said. In the study published in PLoS Genetics, the research team, including co-lead authors Betty M. Booker, PhD, a post-doctoral fellow in Ahituv's laboratory, and Tara Friedrich, a PhD student at UCSF and Gladstone Institutes, searched for the evolutionary origin of the bat wing. "We identified genomic sequences that have not changed in most vertebrates, but experienced rapid changes in the common ancestor of today's bats," explained Friedrich, a member of the laboratory of co-senior investigator Katherine S. Pollard, PhD, a senior investigator at the Gladstone Institutes, a UCSF professor of epidemiology and biostatistics, and a member of the UCSF Institute for Human Genetics. The team mapped these so-called "bat accelerated regions" (BARs) onto areas that were predicted to be important switches that turn genes on during limb development, and found 166 BARs with the potential to influence bat wing development. The researchers tested the effects of five of these BARs in genetically modified mouse embryos and found that all five bat sequences were capable of switching on a reporter gene in the developing mouse forelimb. They noted that one region, BAR116, is located near the HoxD genes, which are known to be involved in limb patterning and skeletal growth. Previously, Mandy Mason, a PhD student at UCT, had shown that two of the HoxD genes—Hoxd10 and Hoxd11—are far more active in bat wings compared to bat legs during their embryonic development. Following up these lines of evidence, the researchers showed that the bat BAR116 sequence appears to function as a genetic switch that is active in developing limbs, in particular the forelimbs, while the equivalent mouse sequence did not show any activity. "Our computational method enabled identification of DNA sequences that changed dramatically during the emergence of bats," said Pollard. "It is exciting to see that this evolutionary signature pointed us to parts of the mammalian genome that control limb development." In addition to unveiling new fundamental details of the evolutionary and developmental origins of powered flight in bats, the new research may provide broader insights into the biological processes that control how mammalian limbs develop in general, Ahituv said. "Importantly, this work will increase our understanding of how alterations in limb development could lead to limb malformations in humans," he said. "Potentially, it could eventually help contribute to the development of tools and techniques to prevent such malformations." More information: Betty M. Booker et al. Bat Accelerated Regions Identify a Bat Forelimb Specific Enhancer in the HoxD Locus, PLOS Genetics (2016). DOI: 10.1371/journal.pgen.1005738 Walter L Eckalbar et al. Transcriptomic and epigenomic characterization of the developing bat wing, Nature Genetics (2016). DOI: 10.1038/ng.3537


News Article | January 27, 2016
Site: www.nature.com

Many games of perfect information, such as chess, checkers, othello, backgammon and Go, may be defined as alternating Markov games39. In these games, there is a state space (where state includes an indication of the current player to play); an action space defining the legal actions in any given state s ∈  ; a state transition function f(s, a, ξ) defining the successor state after selecting action a in state s and random input ξ (for example, dice); and finally a reward function ri(s) describing the reward received by player i in state s. We restrict our attention to two-player zero-sum games, r1(s) = −r2(s) = r(s), with deterministic state transitions, f(s, a, ξ) = f(s, a), and zero rewards except at a terminal time step T. The outcome of the game z  = ±r(s ) is the terminal reward at the end of the game from the perspective of the current player at time step t. A policy p(a|s) is a probability distribution over legal actions . A value function is the expected outcome if all actions for both players are selected according to policy p, that is,  . Zero-sum games have a unique optimal value function v*(s) that determines the outcome from state s following perfect play by both players, The optimal value function can be computed recursively by minimax (or equivalently negamax) search40. Most games are too large for exhaustive minimax tree search; instead, the game is truncated by using an approximate value function v(s) ≈ v*(s) in place of terminal rewards. Depth-first minimax search with alpha–beta pruning40 has achieved superhuman performance in chess4, checkers5 and othello6, but it has not been effective in Go7. Reinforcement learning can learn to approximate the optimal value function directly from games of self-play39. The majority of prior work has focused on a linear combination v (s) = φ(s) · θ of features φ(s) with weights θ. Weights were trained using temporal-difference learning41 in chess42, 43, checkers44, 45 and Go30; or using linear regression in othello6 and Scrabble9. Temporal-difference learning has also been used to train a neural network to approximate the optimal value function, achieving superhuman performance in backgammon46; and achieving weak kyu-level performance in small-board Go28, 29, 47 using convolutional networks. An alternative approach to minimax search is Monte Carlo tree search (MCTS)11, 12, which estimates the optimal value of interior nodes by a double approximation, . The first approximation, , uses n Monte Carlo simulations to estimate the value function of a simulation policy Pn. The second approximation, , uses a simulation policy Pn in place of minimax optimal actions. The simulation policy selects actions according to a search control function , such as UCT12, that selects children with higher action values, Qn(s, a) = −Vn(f(s, a)), plus a bonus u(s, a) that encourages exploration; or in the absence of a search tree at state s, it samples actions from a fast rollout policy  . As more simulations are executed and the search tree grows deeper, the simulation policy becomes informed by increasingly accurate statistics. In the limit, both approximations become exact and MCTS (for example, with UCT) converges12 to the optimal value function . The strongest current Go programs are based on MCTS13, 14, 15, 36. MCTS has previously been combined with a policy that is used to narrow the beam of the search tree to high-probability moves13; or to bias the bonus term towards high-probability moves48. MCTS has also been combined with a value function that is used to initialize action values in newly expanded nodes16, or to mix Monte Carlo evaluation with minimax evaluation49. By contrast, AlphaGo’s use of value functions is based on truncated Monte Carlo search algorithms8, 9, which terminate rollouts before the end of the game and use a value function in place of the terminal reward. AlphaGo’s position evaluation mixes full rollouts with truncated rollouts, resembling in some respects the well-known temporal-difference learning algorithm TD(λ). AlphaGo also differs from prior work by using slower but more powerful representations of the policy and value function; evaluating deep neural networks is several orders of magnitude slower than linear representations and must therefore occur asynchronously. The performance of MCTS is to a large degree determined by the quality of the rollout policy. Prior work has focused on handcrafted patterns50 or learning rollout policies by supervised learning13, reinforcement learning16, simulation balancing51, 52 or online adaptation30, 53; however, it is known that rollout-based position evaluation is frequently inaccurate54. AlphaGo uses relatively simple rollouts, and instead addresses the challenging problem of position evaluation more directly using value networks. To efficiently integrate large neural networks into AlphaGo, we implemented an asynchronous policy and value MCTS algorithm (APV-MCTS). Each node s in the search tree contains edges (s, a) for all legal actions . Each edge stores a set of statistics, where P(s, a) is the prior probability, W (s, a) and W (s, a) are Monte Carlo estimates of total action value, accumulated over N (s, a) and N (s, a) leaf evaluations and rollout rewards, respectively, and Q(s, a) is the combined mean action value for that edge. Multiple simulations are executed in parallel on separate search threads. The APV-MCTS algorithm proceeds in the four stages outlined in Fig. 3. Selection (Fig. 3a). The first in-tree phase of each simulation begins at the root of the search tree and finishes when the simulation reaches a leaf node at time step L. At each of these time steps, t < L, an action is selected according to the statistics in the search tree, using a variant of the PUCT algorithm48, , where c is a constant determining the level of exploration; this search control strategy initially prefers actions with high prior probability and low visit count, but asymptotically prefers actions with high action value. Evaluation (Fig. 3c). The leaf position s is added to a queue for evaluation v (s ) by the value network, unless it has previously been evaluated. The second rollout phase of each simulation begins at leaf node s and continues until the end of the game. At each of these time-steps, t ≥ L, actions are selected by both players according to the rollout policy, . When the game reaches a terminal state, the outcome is computed from the final score. Backup (Fig. 3d). At each in-tree step t ≤ L of the simulation, the rollout statistics are updated as if it has lost n games, N (s , a ) ← N (s , a ) + n ; W (s , a ) ← W (s , a ) −n ; this virtual loss55 discourages other threads from simultaneously exploring the identical variation. At the end of the simulation, t he rollout statistics are updated in a backward pass through each step t ≤ L, replacing the virtual losses by the outcome, N (s , a ) ← N (s , a ) −n  + 1; W (s , a ) ← W (s , a ) + n  + z . Asynchronously, a separate backward pass is initiated when the evaluation of the leaf position s completes. The output of the value network v (s ) is used to update value statistics in a second backward pass through each step t ≤ L, N (s , a ) ← N (s , a ) + 1, W (s , a ) ← W (s , a ) + v (s ). The overall evaluation of each state action is a weighted average of the Monte Carlo estimates, , that mixes together the value network and rollout evaluations with weighting parameter λ. All updates are performed lock-free56. Expansion (Fig. 3b). When the visit count exceeds a threshold, N (s, a) > n , the successor state s′ = f(s, a) is added to the search tree. The new node is initialized to {N (s′, a) = N (s′, a) = 0, W (s′, a) = W (s′, a) = 0, P(s′,a) = p (a|s′)}, using a tree policy p (a|s′) (similar to the rollout policy but with more features, see Extended Data Table 4) to provide placeholder prior probabilities for action selection. The position s′ is also inserted into a queue for asynchronous GPU evaluation by the policy network. Prior probabilities are computed by the SL policy network with a softmax temperature set to β; these replace the placeholder prior probabilities, , using an atomic update. The threshold n is adjusted dynamically to ensure that the rate at which positions are added to the policy queue matches the rate at which the GPUs evaluate the policy network. Positions are evaluated by both the policy network and the value network using a mini-batch size of 1 to minimize end-to-end evaluation time. We also implemented a distributed APV-MCTS algorithm. This architecture consists of a single master machine that executes the main search, many remote worker CPUs that execute asynchronous rollouts, and many remote worker GPUs that execute asynchronous policy and value network evaluations. The entire search tree is stored on the master, which only executes the in-tree phase of each simulation. The leaf positions are communicated to the worker CPUs, which execute the rollout phase of simulation, and to the worker GPUs, which compute network features and evaluate the policy and value networks. The prior probabilities of the policy network are returned to the master, where they replace placeholder prior probabilities at the newly expanded node. The rewards from rollouts and the value network outputs are each returned to the master, and backed up the originating search path. At the end of search AlphaGo selects the action with maximum visit count; this is less sensitive to outliers than maximizing action value15. The search tree is reused at subsequent time steps: the child node corresponding to the played action becomes the new root node; the subtree below this child is retained along with all its statistics, while the remainder of the tree is discarded. The match version of AlphaGo continues searching during the opponent’s move. It extends the search if the action maximizing visit count and the action maximizing action value disagree. Time controls were otherwise shaped to use most time in the middle-game57. AlphaGo resigns when its overall evaluation drops below an estimated 10% probability of winning the game, that is, . AlphaGo does not employ the all-moves-as-first10 or rapid action value estimation58 heuristics used in the majority of Monte Carlo Go programs; when using policy networks as prior knowledge, these biased heuristics do not appear to give any additional benefit. In addition AlphaGo does not use progressive widening13, dynamic komi59 or an opening book60. The parameters used by AlphaGo in the Fan Hui match are listed in Extended Data Table 5. The rollout policy is a linear softmax policy based on fast, incrementally computed, local pattern-based features consisting of both ‘response’ patterns around the previous move that led to state s, and ‘non-response’ patterns around the candidate move a in state s. Each non-response pattern is a binary feature matching a specific 3 × 3 pattern centred on a, defined by the colour (black, white, empty) and liberty count (1, 2, ≥3) for each adjacent intersection. Each response pattern is a binary feature matching the colour and liberty count in a 12-point diamond-shaped pattern21 centred around the previous move. Additionally, a small number of handcrafted local features encode common-sense Go rules (see Extended Data Table 4). Similar to the policy network, the weights π of the rollout policy are trained from 8 million positions from human games on the Tygem server to maximize log likelihood by stochastic gradient descent. Rollouts execute at approximately 1,000 simulations per second per CPU thread on an empty board. Our rollout policy p (a|s) contains less handcrafted knowledge than state-of-the-art Go programs13. Instead, we exploit the higher-quality action selection within MCTS, which is informed both by the search tree and the policy network. We introduce a new technique that caches all moves from the search tree and then plays similar moves during rollouts; a generalization of the ‘last good reply’ heuristic53. At every step of the tree traversal, the most probable action is inserted into a hash table, along with the 3 × 3 pattern context (colour, liberty and stone counts) around both the previous move and the current move. At each step of the rollout, the pattern context is matched against the hash table; if a match is found then the stored move is played with high probability. In previous work, the symmetries of Go have been exploited by using rotationally and reflectionally invariant filters in the convolutional layers24, 28, 29. Although this may be effective in small neural networks, it actually hurts performance in larger networks, as it prevents the intermediate filters from identifying specific asymmetric patterns23. Instead, we exploit symmetries at run-time by dynamically transforming each position s using the dihedral group of eight reflections and rotations, d (s), …, d (s). In an explicit symmetry ensemble, a mini-batch of all 8 positions is passed into the policy network or value network and computed in parallel. For the value network, the output values are simply averaged, . For the policy network, the planes of output probabilities are rotated/reflected back into the original orientation, and averaged together to provide an ensemble prediction, ; this approach was used in our raw network evaluation (see Extended Data Table 3). Instead, APV-MCTS makes use of an implicit symmetry ensemble that randomly selects a single rotation/reflection j ∈ [1, 8] for each evaluation. We compute exactly one evaluation for that orientation only; in each simulation we compute the value of leaf node s by v (d (s )), and allow the search procedure to average over these evaluations. Similarly, we compute the policy network for a single, randomly selected rotation/reflection, . We trained the policy network p to classify positions according to expert moves played in the KGS data set. This data set contains 29.4 million positions from 160,000 games played by KGS 6 to 9 dan human players; 35.4% of the games are handicap games. The data set was split into a test set (the first million positions) and a training set (the remaining 28.4 million positions). Pass moves were excluded from the data set. Each position consisted of a raw board description s and the move a selected by the human. We augmented the data set to include all eight reflections and rotations of each position. Symmetry augmentation and input features were pre-computed for each position. For each training step, we sampled a randomly selected mini-batch of m samples from the augmented KGS data set, and applied an asynchronous stochastic gradient descent update to maximize the log likelihood of the action, . The step size α was initialized to 0.003 and was halved every 80 million training steps, without momentum terms, and a mini-batch size of m = 16. Updates were applied asynchronously on 50 GPUs using DistBelief 61; gradients older than 100 steps were discarded. Training took around 3 weeks for 340 million training steps. We further trained the policy network by policy gradient reinforcement learning25, 26. Each iteration consisted of a mini-batch of n games played in parallel, between the current policy network p that is being trained, and an opponent that uses parameters ρ− from a previous iteration, randomly sampled from a pool of opponents, so as to increase the stability of training. Weights were initialized to ρ = ρ− = σ. Every 500 iterations, we added the current parameters ρ to the opponent pool. Each game i in the mini-batch was played out until termination at step Ti, and then scored to determine the outcome from each player’s perspective. The games were then replayed to determine the policy gradient update, , using the REINFORCE algorithm25 with baseline for variance reduction. On the first pass through the training pipeline, the baseline was set to zero; on the second pass we used the value network v (s) as a baseline; this provided a small performance boost. The policy network was trained in this way for 10,000 mini-batches of 128 games, using 50 GPUs, for one day. We trained a value network to approximate the value function of the RL policy network p . To avoid overfitting to the strongly correlated positions within games, we constructed a new data set of uncorrelated self-play positions. This data set consisted of over 30 million positions, each drawn from a unique game of self-play. Each game was generated in three phases by randomly sampling a time step U ~ unif{1, 450}, and sampling the first t = 1,… U − 1 moves from the SL policy network, a  ~ p (·|s ); then sampling one move uniformly at random from available moves, a  ~ unif{1, 361} (repeatedly until a is legal); then sampling the remaining sequence of moves until the game terminates, t = U + 1, … T, from the RL policy network, a  ~ p (·|s ). Finally, the game is scored to determine the outcome z  = ±r(s ). Only a single training example (s , z ) is added to the data set from each game. This data provides unbiased samples of the value function . During the first two phases of generation we sample from noisier distributions so as to increase the diversity of the data set. The training method was identical to SL policy network training, except that the parameter update was based on mean squared error between the predicted values and the observed rewards, . The value network was trained for 50 million mini-batches of 32 positions, using 50 GPUs, for one week. Each position s was pre-processed into a set of 19 × 19 feature planes. The features that we use come directly from the raw representation of the game rules, indicating the status of each intersection of the Go board: stone colour, liberties (adjacent empty points of stone’s chain), captures, legality, turns since stone was played, and (for the value network only) the current colour to play. In addition, we use one simple tactical feature that computes the outcome of a ladder search7. All features were computed relative to the current colour to play; for example, the stone colour at each intersection was represented as either player or opponent rather than black or white. Each integer feature value is split into multiple 19 × 19 planes of binary values (one-hot encoding). For example, separate binary feature planes are used to represent whether an intersection has 1 liberty, 2 liberties,…, ≥8 liberties. The full set of feature planes are listed in Extended Data Table 2. The input to the policy network is a 19 × 19 × 48 image stack consisting of 48 feature planes. The first hidden layer zero pads the input into a 23 × 23 image, then convolves k filters of kernel size 5 × 5 with stride 1 with the input image and applies a rectifier nonlinearity. Each of the subsequent hidden layers 2 to 12 zero pads the respective previous hidden layer into a 21 × 21 image, then convolves k filters of kernel size 3 × 3 with stride 1, again followed by a rectifier nonlinearity. The final layer convolves 1 filter of kernel size 1 × 1 with stride 1, with a different bias for each position, and applies a softmax function. The match version of AlphaGo used k = 192 filters; Fig. 2b and Extended Data Table 3 additionally show the results of training with k = 128, 256 and 384 filters. The input to the value network is also a 19 × 19 × 48 image stack, with an additional binary feature plane describing the current colour to play. Hidden layers 2 to 11 are identical to the policy network, hidden layer 12 is an additional convolution layer, hidden layer 13 convolves 1 filter of kernel size 1 × 1 with stride 1, and hidden layer 14 is a fully connected linear layer with 256 rectifier units. The output layer is a fully connected linear layer with a single tanh unit. We evaluated the relative strength of computer Go programs by running an internal tournament and measuring the Elo rating of each program. We estimate the probability that program a will beat program b by a logistic function , and estimate the ratings e(·) by Bayesian logistic regression, computed by the BayesElo program37 using the standard constant c  = 1/400. The scale was anchored to the BayesElo rating of professional Go player Fan Hui (2,908 at date of submission)62. All programs received a maximum of 5 s computation time per move; games were scored using Chinese rules with a komi of 7.5 points (extra points to compensate white for playing second). We also played handicap games where AlphaGo played white against existing Go programs; for these games we used a non-standard handicap system in which komi was retained but black was given additional stones on the usual handicap points. Using these rules, a handicap of K stones is equivalent to giving K − 1 free moves to black, rather than K − 1/2 free moves using standard no-komi handicap rules. We used these handicap rules because AlphaGo’s value network was trained specifically to use a komi of 7.5. With the exception of distributed AlphaGo, each computer Go program was executed on its own single machine, with identical specifications, using the latest available version and the best hardware configuration supported by that program (see Extended Data Table 6). In Fig. 4, approximate ranks of computer programs are based on the highest KGS rank achieved by that program; however, the KGS version may differ from the publicly available version. The match against Fan Hui was arbitrated by an impartial referee. Five formal games and five informal games were played with 7.5 komi, no handicap, and Chinese rules. AlphaGo won these games 5–0 and 3–2 respectively (Fig. 6 and Extended Data Table 1). Time controls for formal games were 1 h main time plus three periods of 30 s byoyomi. Time controls for informal games were three periods of 30 s byoyomi. Time controls and playing conditions were chosen by Fan Hui in advance of the match; it was also agreed that the overall match outcome would be determined solely by the formal games. To approximately assess the relative rating of Fan Hui to computer Go programs, we appended the results of all ten games to our internal tournament results, ignoring differences in time controls.


News Article | January 20, 2016
Site: phys.org

The crimson stigma of the saffron flower (Crocus sativus) is one of the oldest and most expensive spices in the world, particularly those varieties which are internationally recognised for their quality, such as saffron grown in Spain. This has led to the fraudulent labelling of non-Spanish saffron. "Over the past few years the media have been reporting this fraudulent activity, but up until now there were barely any analytical tools that could be used to detect said fraud. So, we created a new strategy to determine the authenticity of saffron based on metabolomics or, in other words, the chemical fingerprints of foods," explains Josep Rubert, a researcher at University of Chemistry and Technology (UCT Prague, Czech Republic) and the University of Valencia (Spain). The new technique allows for three types of saffron to be defined: one which is certified with the Protected Designation of Origin (PDO) from La Mancha or Aragon, another which is grown and packaged in Spain (although it does not have the PDO certificate) and a third category which is packaged as 'Spanish saffron' but, despite its name, is of unknown origin (although most likely packaged in Spain). With these possibilities, scientists from the UCT Prague leaded by Prof. Jana Hajslova—and where Rubert is also carrying out postdoctoral research including this study —, collected 44 commercial saffron samples in order to test the authenticity of what's stated in the product labels. The findings, published this month by the journal Food Chemistry, revealed that more than 50% of the samples were fraudulent, as 26 ones labelled as 'Spanish saffron' were neither grown nor processed in Spain. "It is highly likely that lower quality saffron is purchased in other countries (such as Morocoo, Iran and India according to our data) at a much lower price than in Spain -indicates the researcher —, to later be packaged and sold as Spanish saffron despite being of unknown origin a fraudulent activity that gambles with consumers' trust". The technique developed by scientists from the Czech Republic and Spain has confirmed that the saffron labelled with the PDO Certificate from La Mancha (and Aragon) were indeed grown and processed in Spain. "Here there was no fraudulent activity the saffron perfectly matched up with our models," emphasises Rubert, "unlike the samples of 'Spanish saffron' that had either a completely different chemical fingerprint or a different collection of small molecules". Chemistry and statistics to expose the fraud The authors of this study combined chemistry with statistics in order to develop their methodology. The first phase of the study consisted in identifying the metabolites or small molecules characteristic of saffron. After, a method was created to detect these small molecules using liquid chromatography coupled with high-resolution mass spectrometry. On one hand, the statistical analyses have served to detect the clear differences between the three types of saffron in addition to validating the technique. According to the authors, the result "is a top-quality model that correctly classified 100% of these samples in addition to having the capacity to correctly categorise others (even if they are unknown and do not have a label) more than 85% of the time". The authors suggest that glycerophospholipids and their oxidised lipids are the best molecular markers for determining the origin of saffron. They have also observed that the saffron technology and processing play a crucial role, "specifically during the drying process, wherein transformation of the product is determined by the temperature at which the process is carried out. The place where the saffron originates also has an influence on the end product". For saffron originating from La Mancha, for example, the drying process involves laying out the fresh stigmas over sieves that are placed next to a heat source such as a fire, hot coal, a stove or a brazier. Saffron dehydration happens quickly -in half an hour- and is carried out at a temperature of 70 ºC which accelerates lipid oxidation. Over recent decades, saffron originating from Castile-La Mancha has represented over 97% of Spain's domestic production a statistic that presents an enormous gap with regard to exportations. Between 1997 and 2013, an average of 2,813 kg of saffron was produced annually in Spain. However, Spain exported 35,978 kg of this product on average each year. Where did those remaining 33,165 kg come from? "They came from other countries, such as Iran or Morocco," mentions Pedro M. Pérez again, manager of the Protected Designation of Origin Regulatory Body in La Mancha. He insists that: "That foreign saffron is brought to Spain and labelled as 'produced and packaged in Spain', which is true, but the label fails to indicate the saffron's true origin, meaning that the consumer does not have enough information to assess the product". The manager of the regulatory body reiterates that there is a Spanish national law dated 1999 in addition to a European law from 2011 regarding the proper labelling of foodstuffs, "but the competent authorities of Spain's Autonomous Communities are not successfully fulfilling their responsibilities with regard to saffron". Explore further: Smartphone maker HTC invests in UK, US firms More information: Josep Rubert et al. Saffron authentication based on liquid chromatography high resolution tandem mass spectrometry and multivariate data analysis, Food Chemistry (2016). DOI: 10.1016/j.foodchem.2016.01.003


A notification appliance circuit (NAC) includes notification devices having a high capacity rechargeable energy storage device such as a supercapacitor and a strobe circuit. The supercapacitor can provide energy to produce flashes over an extended time period without fully discharging. The notification devices can also make use of the fallback power strategy in which the strobe circuit operates with reduced intensity while the supercapacitor is being recharged.

Loading UCT collaborators
Loading UCT collaborators