Ageing is a fundamental, unsolved mystery in biology. DAF-16, a FOXO-family transcription factor, influences the rate of ageing of Caenorhabditis elegans in response to insulin/insulin-like growth factor 1 (IGF-I) signalling. Using DNA microarray analysis, we have found that DAF-16 affects expression of a set of genes during early adulthood, the time at which this pathway is known to control ageing. Here we find that many of these genes influence the ageing process. The insulin/IGF-I pathway functions cell non-autonomously to regulate lifespan, and our findings suggest that it signals other cells, at least in part, by feedback regulation of an insulin/IGF-I homologue. Furthermore, our findings suggest that the insulin/IGF-I pathway ultimately exerts its effect on lifespan by upregulating a wide variety of genes, including cellular stress-response, antimicrobial and metabolic genes, and by downregulating specific life-shortening genes.

Artificial intelligence (AI) aims to mimic human cognitive functions. It is bringing a paradigm shift to healthcare, powered by increasing availability of healthcare data and rapid progress of analytics techniques. We survey the current status of AI applications in healthcare and discuss its future. AI can be applied to various types of healthcare data (structured and unstructured). Popular AI techniques include machine learning methods for structured data, such as the classical support vector machine and neural network, and the modern deep learning, as well as natural language processing for unstructured data. Major disease areas that use AI tools include cancer, neurology and cardiology. We then review in more details the AI applications in stroke, in the three major areas of early detection and diagnosis, treatment, as well as outcome prediction and prognosis evaluation. We conclude with discussion about pioneer AI systems, such as IBM Watson, and hurdles for real-life deployment of AI.

The pancreas contains both exocrine and endocrine cells, but the molecular mechanisms controlling the differentiation of these cell types are largely unknown. Despite their endodermal origin, pancreatic endocrine cells share several molecular characteristics with neurons, and, like neurons in the central nervous system, differentiating endocrine cells in the pancreas appear in a scattered fashion within a field of progenitor cells. This indicates that they may be generated by lateral specification through Notch signalling. Here, to test this idea, we analysed pancreas development in mice genetically altered at several steps in the Notch signalling pathway. Mice deficient for Delta-like gene 1 (Dll1) or the intracellular mediator RBP-Jkappa showed accelerated differentiation of pancreatic endocrine cells. A similar phenotype was observed in mice over-expressing neurogenin 3 (ngn 3) or the intracellular form of Notch3 (a repressor of Notch signalling). These data provide evidence that ngn3 acts as proendocrine gene and that Notch signalling is critical for the decision between the endocrine and progenitor/exocrine fates in the developing pancreas.

Even though the well-established Haber-Bosch process has been the major artificial way to "fertilize" the earth, its energy-intensive nature has been motivating people to learn from nitrogenase, which can fix atmospheric N2 to NH3 in vivo under mild conditions with its precisely arranged proteins. Here we demonstrate that efficient fixation of N2 to NH3 can proceed under room temperature and atmospheric pressure in water using visible light illuminated BiOBr nanosheets of oxygen vacancies in the absence of any organic scavengers and precious-metal cocatalysts. The designed catalytic oxygen vacancies of BiOBr nanosheets on the exposed {001} facets, with the availability of localized electrons for π-back-donation, have the ability to activate the adsorbed N2, which can thus be efficiently reduced to NH3 by the interfacial electrons transferred from the excited BiOBr nanosheets. This study might open up a new vista to fix atmospheric N2 to NH3 through the less energy-demanding photochemical process.

Among patients with TIA or minor stroke who can be treated within 24 hours after the onset of symptoms, the combination of clopidogrel and aspirin is superior to aspirin alone for reducing the risk of stroke in the first 90 days and does not increase the risk of hemorrhage. (Funded by the Ministry of Science and Technology of the People's Republic of China; CHANCE ClinicalTrials.gov number, NCT00979589.).

We report a surface segregation approach to synthesize high quality graphenes on Ni under ambient pressure. Graphenes were segregated from Ni surfaces by carbon dissolving at high temperature and cooling down with various cooling rates. Different segregation behaviors were identified, allowing us to control the thickness and defects of graphene films. Electron microscopy and Raman spectroscopy studies indicated that these graphenes have high quality crystalline structure and controllable thickness. Graphenes were transferred to insulating substrates by wet etching and were found to maintain their high quality.

The success of CNNs in various applications is accompanied by a significant increase in the computation and parameter storage costs. Recent efforts toward reducing these overheads involve pruning and compressing the weights of various layers without hurting original accuracy. However, magnitude-based pruning of weights reduces a significant number of parameters from the fully connected layers and may not adequately reduce the computation costs in the convolutional layers due to irregular sparsity in the pruned networks. We present an acceleration method for CNNs, where we prune filters from CNNs that are identified as having a small effect on the output accuracy. By removing whole filters in the network together with their connecting feature maps, the computation costs are reduced significantly. In contrast to pruning weights, this approach does not result in sparse connectivity patterns. Hence, it does not need the support of sparse convolution libraries and can work with existing efficient BLAS libraries for dense matrix multiplications. We show that even simple filter pruning techniques can reduce inference costs for VGG-16 by up to 34% and ResNet-110 by up to 38% on CIFAR10 while regaining close to the original accuracy by retraining the networks.

Protein structures in nature often exhibit a high degree of regularity (for example, secondary structure and tertiary symmetries) that is absent from random compact conformations. With the use of a simple lattice model of protein folding, it was demonstrated that structural regularities are related to high "designability" and evolutionary stability. The designability of each compact structure is measured by the number of sequences that can design the structure-that is, sequences that possess the structure as their nondegenerate ground state. Compact structures differ markedly in terms of their designability; highly designable structures emerge with a number of associated sequences much larger than the average. These highly designable structures possess "proteinlike" secondary structure and even tertiary symmetries. In addition, they are thermodynamically more stable than other structures. These results suggest that protein structures are selected in nature because they are readily designed and stable against mutations, and that such a selection simultaneously leads to thermodynamic stability.