Abstract: We explore in a parameterized manner a very large range of physically plausible equations of state (EOSs) for compact stars for matter that is either purely hadronic or that exhibits a phase transition. In particular, we produce two classes of EOSs with and without phase transitions, each containing one million EOSs. We then impose constraints on the maximum mass ($Ml2.16\text{ }\text{ }{M}_{\ensuremath{\bigodot}}$) and on the dimensionless tidal deformability ($\stackrel{\texttildelow{}}{\mathrm{\ensuremath{\Lambda}}}l800$) deduced from GW170817, together with recent suggestions of lower limits on $\stackrel{\texttildelow{}}{\mathrm{\ensuremath{\Lambda}}}$. Exploiting more than $1{0}^{9}$ equilibrium models for each class of EOSs, we produce distribution functions of all the stellar properties and determine, among other quantities, the radius that is statistically most probable for any value of the stellar mass. In this way, we deduce that the radius of a purely hadronic neutron star with a representative mass of $1.4\text{ }\text{ }{M}_{\ensuremath{\bigodot}}$ is constrained to be $12.00l{R}_{1.4}/\mathrm{km}l13.45$ at a $2\ensuremath{\sigma}$ confidence level, with a most likely value of ${\overline{R}}_{1.4}=12.39\text{ }\text{ }\mathrm{km}$; similarly, the smallest dimensionless tidal deformability is ${\stackrel{\texttildelow{}}{\mathrm{\ensuremath{\Lambda}}}}_{1.4}g375$, again at a $2\ensuremath{\sigma}$ level. On the other hand, because EOSs with a phase transition allow for very compact stars on the so-called ``twin-star'' branch, small radii are possible with such EOSs although not probable, i.e., $8.53l{R}_{1.4}/\mathrm{km}l13.74$ and ${\overline{R}}_{1.4}=13.06\text{ }\text{ }\mathrm{km}$ at a $2\ensuremath{\sigma}$ level, with ${\stackrel{\texttildelow{}}{\mathrm{\ensuremath{\Lambda}}}}_{1.4}g35.5$ at a $3\ensuremath{\sigma}$ level. Finally, since these EOSs exhibit upper limits on $\stackrel{\texttildelow{}}{\mathrm{\ensuremath{\Lambda}}}$, the detection of a binary with a total mass of $3.4\text{ }\text{ }{M}_{\ensuremath{\bigodot}}$ and ${\stackrel{\texttildelow{}}{\mathrm{\ensuremath{\Lambda}}}}_{1.7}g461$ can rule out twin-star solutions.

Abstract: We present diffusion-convolutional neural networks (DCNNs), a new model for graph-structured data. Through the introduction of a diffusion-convolution operation, we show how diffusion-based representations can be learned from graph-structured data and used as an effective basis for node classification. DCNNs have several attractive qualities, including a latent representation for graphical data that is invariant under isomorphism, as well as polynomial-time prediction and learning that can be represented as tensor operations and efficiently implemented on a GPU. Through several experiments with real structured datasets, we demonstrate that DCNNs are able to outperform probabilistic relational models and kernel-on-graph methods at relational node classification tasks.

Abstract: The past 20 years have witnessed unprecedented progress in brain-computer interfaces (BCIs). However, low communication rates remain key obstacles to BCI-based communication in humans. This study presents an electroencephalogram-based BCI speller that can achieve information transfer rates (ITRs) up to 5.32 bits per second, the highest ITRs reported in BCI spellers using either noninvasive or invasive methods. Based on extremely high consistency of frequency and phase observed between visual flickering signals and the elicited single-trial steady-state visual evoked potentials, this study developed a synchronous modulation and demodulation paradigm to implement the speller. Specifically, this study proposed a new joint frequency-phase modulation method to tag 40 characters with 0.5-s-long flickering signals and developed a user-specific target identification algorithm using individual calibration data. The speller achieved high ITRs in online spelling tasks. This study demonstrates that BCIs can provide a truly naturalistic high-speed communication channel using noninvasively recorded brain activities.

Abstract: Fossil evidence points to an African origin of Homo sapiens from a group called either H. heidelbergensis or H. rhodesiensis. However, the exact place and time of emergence of H. sapiens remain obscure because the fossil record is scarce and the chronological age of many key specimens remains uncertain. In particular, it is unclear whether the present day ‘modern’ morphology rapidly emerged approximately 200 thousand years ago (ka) among earlier representatives of H. sapiens1 or evolved gradually over the last 400 thousand years2. Here we report newly discovered human fossils from Jebel Irhoud, Morocco, and interpret the affinities of the hominins from this site with other archaic and recent human groups. We identified a mosaic of features including facial, mandibular and dental morphology that aligns the Jebel Irhoud material with early or recent anatomically modern humans and more primitive neurocranial and endocranial morphology. In combination with an age of 315 ± 34 thousand years (as determined by thermoluminescence dating)3, this evidence makes Jebel Irhoud the oldest and richest African Middle Stone Age hominin site that documents early stages of the H. sapiens clade in which key features of modern morphology were established. Furthermore, it shows that the evolutionary processes behind the emergence of H. sapiens involved the whole African continent.

Abstract: In the future, artificial intelligence (AI) is likely to substantially change both marketing strategies and customer behaviors. Building from not only extant research but also extensive interactions with practice, the authors propose a multidimensional framework for understanding the impact of AI involving intelligence levels, task types, and whether AI is embedded in a robot. Prior research typically addresses a subset of these dimensions; this paper integrates all three into a single framework. Next, the authors propose a research agenda that addresses not only how marketing strategies and customer behaviors will change in the future, but also highlights important policy questions relating to privacy, bias and ethics. Finally, the authors suggest AI will be more effective if it augments (rather than replaces) human managers.

Abstract: During the last decade, Convolutional Neural Networks (CNNs) have become the de facto standard for various Computer Vision and Machine Learning operations. CNNs are feed-forward Artificial Neural Networks (ANNs) with alternating convolutional and subsampling layers. Deep 2D CNNs with many hidden layers and millions of parameters have the ability to learn complex objects and patterns providing that they can be trained on a massive size visual database with ground-truth labels. With a proper training, this unique ability makes them the primary tool for various engineering applications for 2D signals such as images and video frames. Yet, this may not be a viable option in numerous applications over 1D signals especially when the training data is scarce or application-specific. To address this issue, 1D CNNs have recently been proposed and immediately achieved the state-of-the-art performance levels in several applications such as personalized biomedical data classification and early diagnosis, structural health monitoring, anomaly detection and identification in power electronics and motor-fault detection. Another major advantage is that a real-time and low-cost hardware implementation is feasible due to the simple and compact configuration of 1D CNNs that perform only 1D convolutions (scalar multiplications and additions). This paper presents a comprehensive review of the general architecture and principals of 1D CNNs along with their major engineering applications, especially focused on the recent progress in this field. Their state-of-the-art performance is highlighted concluding with their unique properties. The benchmark datasets and the principal 1D CNN software used in those applications are also publically shared in a dedicated website.

Abstract: In the growth condition(s) of plants, numerous secondary metabolites (SMs) are produced by them to serve variety of cellular functions essential for physiological processes, and recent increasing evidences have implicated stress and defense response signaling in their production. The type and concentration(s) of secondary molecule(s) produced by a plant are determined by the species, genotype, physiology, developmental stage and environmental factors during growth. This suggests the physiological adaptive responses employed by various plant taxonomic groups in coping with the stress and defensive stimuli. The past recent decades had witnessed renewed interest to study abiotic factors that influence secondary metabolism during in vitro and in vivo growth of plants. Application of molecular biology tools and techniques are facilitating understanding the signaling processes and pathways involved in the SMs production at subcellular, cellular, organ and whole plant systems during in vivo and in vitro growth, with application in metabolic engineering of biosynthetic pathways intermediates.

Abstract: Although humans have cospeciated with their gut-resident microbes, it is difficult to infer features of our ancestral microbiome. Here, we examine the microbiome profile of 350 stool samples collected longitudinally for more than a year from the Hadza hunter-gatherers of Tanzania. The data reveal annual cyclic reconfiguration of the microbiome, in which some taxa become undetectable only to reappear in a subsequent season. Comparison of the Hadza data set with data collected from 18 populations in 16 countries with varying lifestyles reveals that gut community membership corresponds to modernization: Notably, the taxa within the Hadza that are the most seasonally volatile similarly differentiate industrialized and traditional populations. These data indicate that some dynamic lineages of microbes have decreased in prevalence and abundance in modernized populations.

Abstract: The functional maturation and preservation of hepatic cells derived from human induced pluripotent stem cells (hiPSCs) are essential to personalized in vitro drug screening and disease study. Major liver functions are tightly linked to the 3D assembly of hepatocytes, with the supporting cell types from both endodermal and mesodermal origins in a hexagonal lobule unit. Although there are many reports on functional 2D cell differentiation, few studies have demonstrated the in vitro maturation of hiPSC-derived hepatic progenitor cells (hiPSC-HPCs) in a 3D environment that depicts the physiologically relevant cell combination and microarchitecture. The application of rapid, digital 3D bioprinting to tissue engineering has allowed 3D patterning of multiple cell types in a predefined biomimetic manner. Here we present a 3D hydrogel-based triculture model that embeds hiPSC-HPCs with human umbilical vein endothelial cells and adipose-derived stem cells in a microscale hexagonal architecture. In comparison with 2D monolayer culture and a 3D HPC-only model, our 3D triculture model shows both phenotypic and functional enhancements in the hiPSC-HPCs over weeks of in vitro culture. Specifically, we find improved morphological organization, higher liver-specific gene expression levels, increased metabolic product secretion, and enhanced cytochrome P450 induction. The application of bioprinting technology in tissue engineering enables the development of a 3D biomimetic liver model that recapitulates the native liver module architecture and could be used for various applications such as early drug screening and disease modeling.

Abstract: Neurodegeneration is a phenomenon that occurs in the central nervous system through the hallmarks associating the loss of neuronal structure and function. Neurodegeneration is observed after viral insult and mostly in various so-called 'neurodegenerative diseases', generally observed in the elderly, such as Alzheimer's disease, multiple sclerosis, Parkinson's disease and amyotrophic lateral sclerosis that negatively affect mental and physical functioning. Causative agents of neurodegeneration have yet to be identified. However, recent data have identified the inflammatory process as being closely linked with multiple neurodegenerative pathways, which are associated with depression, a consequence of neurodegenerative disease. Accordingly, pro‑inflammatory cytokines are important in the pathophysiology of depression and dementia. These data suggest that the role of neuroinflammation in neurodegeneration must be fully elucidated, since pro‑inflammatory agents, which are the causative effects of neuroinflammation, occur widely, particularly in the elderly in whom inflammatory mechanisms are linked to the pathogenesis of functional and mental impairments. In this review, we investigated the role played by the inflammatory process in neurodegenerative diseases.

Abstract: Carnosine, homocarnosine, and anserine are present in high concentrations in the muscle and brain of many animals and humans. However, their exact function is not clear. The antioxidant activity of these compounds has been examined by testing their peroxyl radical-trapping ability at physiological concentrations. Carnosine, homocarnosine, anserine, and other histidine derivatives all showed antioxidant activity. All of these compounds showing peroxyl radical-trapping activity were also electrochemically active as reducing agents in cyclic voltammetric measurements. Furthermore, carnosine inhibited the oxidative hydroxylation of deoxyguanosine induced by ascorbic acid and copper ions. Other roles of carnosine, such as chelation of metal ions, quenching of singlet oxygen, and binding of hydroperoxides, are also discussed. The data suggest a role for these histidine-related compounds as endogenous antioxidants in brain and muscle.

Abstract: For the past 15 years, the UCSC Genome Browser (http://genome.ucsc.edu/) has served the international research community by offering an integrated platform for viewing and analyzing information from a large database of genome assemblies and their associated annotations. The UCSC Genome Browser has been under continuous development since its inception with new data sets and software features added frequently. Some release highlights of this year include new and updated genome browsers for various assemblies, including bonobo and zebrafish; new gene annotation sets; improvements to track and assembly hub support; and a new interactive tool, the "Data Integrator", for intersecting data from multiple tracks. We have greatly expanded the data sets available on the most recent human assembly, hg38/GRCh38, to include updated gene prediction sets from GENCODE, more phenotype- and disease-associated variants from ClinVar and ClinGen, more genomic regulatory data, and a new multiple genome alignment.

Abstract: Machine learning has become a pivotal tool for many projects in computational biology, bioinformatics, and health informatics. Nevertheless, beginners and biomedical researchers often do not have enough experience to run a data mining project effectively, and therefore can follow incorrect practices, that may lead to common mistakes or over-optimistic results. With this review, we present ten quick tips to take advantage of machine learning in any computational biology context, by avoiding some common errors that we observed hundreds of times in multiple bioinformatics projects. We believe our ten suggestions can strongly help any machine learning practitioner to carry on a successful project in computational biology and related sciences.

Abstract: Health and fitness applications (apps) have gained popularity in interventions to improve diet, physical activity and sedentary behaviours but their efficacy is unclear. This systematic review examined the efficacy of interventions that use apps to improve diet, physical activity and sedentary behaviour in children and adults. Systematic literature searches were conducted in five databases to identify papers published between 2006 and 2016. Studies were included if they used a smartphone app in an intervention to improve diet, physical activity and/or sedentary behaviour for prevention. Interventions could be stand-alone interventions using an app only, or multi-component interventions including an app as one of several intervention components. Outcomes measured were changes in the health behaviours and related health outcomes (i.e., fitness, body weight, blood pressure, glucose, cholesterol, quality of life). Study inclusion and methodological quality were independently assessed by two reviewers. Twenty-seven studies were included, most were randomised controlled trials (n = 19; 70%). Twenty-three studies targeted adults (17 showed significant health improvements) and four studies targeted children (two demonstrated significant health improvements). Twenty-one studies targeted physical activity (14 showed significant health improvements), 13 studies targeted diet (seven showed significant health improvements) and five studies targeted sedentary behaviour (two showed significant health improvements). More studies (n = 12; 63%) of those reporting significant effects detected between-group improvements in the health behaviour or related health outcomes, whilst fewer studies (n = 8; 42%) reported significant within-group improvements. A larger proportion of multi-component interventions (8 out of 13; 62%) showed significant between-group improvements compared to stand-alone app interventions (5 out of 14; 36%). Eleven studies reported app usage statistics, and three of them demonstrated that higher app usage was associated with improved health outcomes. This review provided modest evidence that app-based interventions to improve diet, physical activity and sedentary behaviours can be effective. Multi-component interventions appear to be more effective than stand-alone app interventions, however, this remains to be confirmed in controlled trials. Future research is needed on the optimal number and combination of app features, behaviour change techniques, and level of participant contact needed to maximise user engagement and intervention efficacy.

Abstract: The precise localization of the repeating fast radio burst (FRB 121102) has provided the first unambiguous association (chance coincidence probability p ≲ 3 × 10‑4) of an FRB with an optical and persistent radio counterpart. We report on optical imaging and spectroscopy of the counterpart and find that it is an extended (0.″6–0.″8) object displaying prominent Balmer and [Oiii] emission lines. Based on the spectrum and emission line ratios, we classify the counterpart as a low-metallicity, star-forming, mr‧ = 25.1 AB mag dwarf galaxy at a redshift of z =0.19273(8), corresponding to a luminosity distance of 972 Mpc. From the angular size, the redshift, and luminosity, we estimate the host galaxy to have a diameter ≲4 kpc and a stellar mass of M* ∼ (4–7) × 107 M⊙, assuming a mass-to-light ratio between 2 to 3 M⊙L⊙‑1. Based on the Hα flux, we estimate the star formation rate of the host to be 0.4 M⊙yr‑1 and a substantial host dispersion measure (DM)depth ≲324 pc cm‑3. The net DM contribution of the host galaxy to FRB 121102 is likely to be lower than this value depending on geometrical factors. We show that the persistent radio source at FRB 121102’s location reported by Marcote et al. is offset from the galaxy’s center of light by ∼200 mas and the host galaxy does not show optical signatures for AGN activity. If FRB121102 is typical of the wider FRB population and if futureinterferometric localizations preferentially find them in dwarf galaxies with low metallicities and prominent emission lines, they would share such a preference with long gamma-ray bursts and superluminous supernovae.

Abstract: Sudden, large-scale and diffuse human migration can amplify localized outbreaks of disease into widespread epidemics1–4. Rapid and accurate tracking of aggregate population flows may therefore be epidemiologically informative. Here we use 11,478,484 counts of mobile phone data from individuals leaving or transiting through the prefecture of Wuhan between 1 January and 24 January 2020 as they moved to 296 prefectures throughout mainland China. First, we document the efficacy of quarantine in ceasing movement. Second, we show that the distribution of population outflow from Wuhan accurately predicts the relative frequency and geographical distribution of infections with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) until 19 February 2020, across mainland China. Third, we develop a spatio-temporal ‘risk source’ model that leverages population flow data (which operationalize the risk that emanates from epidemic epicentres) not only to forecast the distribution of confirmed cases, but also to identify regions that have a high risk of transmission at an early stage. Fourth, we use this risk source model to statistically derive the geographical spread of COVID-19 and the growth pattern based on the population outflow from Wuhan; the model yields a benchmark trend and an index for assessing the risk of community transmission of COVID-19 over time for different locations. This approach can be used by policy-makers in any nation with available data to make rapid and accurate risk assessments and to plan the allocation of limited resources ahead of ongoing outbreaks. Modelling of population flows in China enables the forecasting of the distribution of confirmed cases of COVID-19 and the identification of areas at high risk of SARS-CoV-2 transmission at an early stage.

Abstract: quanteda is an R package providing a comprehensive workflow and toolkit for natural language processing tasks such as corpus management, tokenization, analysis, and visualization. It has extensive functions for applying dictionary analysis, exploring texts using keywords-in-context, computing document and feature similarities, and discovering multi-word expressions through collocation scoring. Based entirely on sparse operations,it provides highly efficient methods for compiling document-feature matrices and for manipulating these or using them in further quantitative analysis. Using C++ and multi-threading extensively, quanteda is also considerably faster and more efficient than other R and Python packages in processing large textual data. The package is designed for R users needing to apply natural language processing to texts,from documents to final analysis. Its capabilities match or exceed those provided in many end-user software applications, many of which are expensive and not open source. The package is therefore of great benefit to researchers, students, and other analysts with fewer financial resources. While using quanteda requires R programming knowledge, its API is designed to enable powerful, efficient analysis with a minimum of steps. By emphasizing consistent design, furthermore, quanteda lowers the barriers to learning and using NLP and quantitative text analysis even for proficient R programmers.

Abstract: Oxygen-derived free radicals and related oxidants are ubiquitous and short-lived intermediates formed in aerobic organisms throughout life. These reactive species participate in redox reactions leading to oxidative modifications in biomolecules, among which proteins and lipids are preferential targets. Despite a broad array of enzymatic and nonenzymatic antioxidant systems in mammalian cells and microbes, excess oxidant formation causes accumulation of new products that may compromise cell function and structure leading to cell degeneration and death. Oxidative events are associated with pathological conditions and the process of normal aging. Notably, physiological levels of oxidants also modulate cellular functions via homeostatic redox-sensitive cell signaling cascades. On the other hand, nitric oxide ( • NO), a free radical and weak oxidant, represents a master physiological regulator via reversible interactions with heme proteins. The bioavailability and actions of • NO are modulated by its fast reaction with superoxide radical ( O 2 • − ), which yields an unusual and reactive peroxide, peroxynitrite, representing the merging of the oxygen radicals and • NO pathways. In this Inaugural Article, I summarize early and remarkable developments in free radical biochemistry and the later evolution of the field toward molecular medicine; this transition includes our contributions disclosing the relationship of • NO with redox intermediates and metabolism. The biochemical characterization, identification, and quantitation of peroxynitrite and its role in disease processes have concentrated much of our attention. Being a mediator of protein oxidation and nitration, lipid peroxidation, mitochondrial dysfunction, and cell death, peroxynitrite represents both a pathophysiologically relevant endogenous cytotoxin and a cytotoxic effector against invading pathogens.

Abstract: Background Ibrutinib has been approved by the Food and Drug Administration for the treatment of patients with untreated chronic lymphocytic leukemia (CLL) since 2016 but has not been compared with chemoimmunotherapy. We conducted a phase 3 trial to evaluate the efficacy of ibrutinib, either alone or in combination with rituximab, relative to chemoimmunotherapy. Methods Patients 65 years of age or older who had untreated CLL were randomly assigned to receive bendamustine plus rituximab, ibrutinib, or ibrutinib plus rituximab. The primary end point was progression-free survival. The Alliance Data and Safety Monitoring Board made the decision to release the data after the protocol-specified efficacy threshold had been met. Results A total of 183 patients were assigned to receive bendamustine plus rituximab, 182 to receive ibrutinib, and 182 to receive ibrutinib plus rituximab. Median progression-free survival was reached only with bendamustine plus rituximab. The estimated percentage of patients wit...

Abstract: A glucose-responsive “closed-loop” insulin delivery system mimicking the function of pancreatic cells has tremendous potential to improve quality of life and health in diabetics. Here, we report a novel glucose-responsive insulin delivery device using a painless microneedle-array patch (“smart insulin patch”) containing glucose-responsive vesicles (GRVs; with an average diameter of 118 nm), which are loaded with insulin and glucose oxidase (GO x ) enzyme. The GRVs are self-assembled from hypoxia-sensitive hyaluronic acid (HS-HA) conjugated with 2-nitroimidazole (NI), a hydrophobic component that can be converted to hydrophilic 2-aminoimidazoles through bioreduction under hypoxic conditions. The local hypoxic microenvironment caused by the enzymatic oxidation of glucose in the hyperglycemic state promotes the reduction of HS-HA, which rapidly triggers the dissociation of vesicles and subsequent release of insulin. The smart insulin patch effectively regulated the blood glucose in a mouse model of chemically induced type 1 diabetes. The described work is the first demonstration, to our knowledge, of a synthetic glucose-responsive device using a hypoxia trigger for regulation of insulin release. The faster responsiveness of this approach holds promise in avoiding hyperglycemia and hypoglycemia if translated for human therapy.

Abstract: Objective To establish whether there is any change in mortality from infection with a new variant of SARS-CoV-2, designated a variant of concern (VOC-202012/1) in December 2020, compared with circulating SARS-CoV-2 variants. Design Matched cohort study. Setting Community based (pillar 2) covid-19 testing centres in the UK using the TaqPath assay (a proxy measure of VOC-202012/1 infection). Participants 54 906 matched pairs of participants who tested positive for SARS-CoV-2 in pillar 2 between 1 October 2020 and 29 January 2021, followed-up until 12 February 2021. Participants were matched on age, sex, ethnicity, index of multiple deprivation, lower tier local authority region, and sample date of positive specimens, and differed only by detectability of the spike protein gene using the TaqPath assay. Main outcome measure Death within 28 days of the first positive SARS-CoV-2 test result. Results The mortality hazard ratio associated with infection with VOC-202012/1 compared with infection with previously circulating variants was 1.64 (95% confidence interval 1.32 to 2.04) in patients who tested positive for covid-19 in the community. In this comparatively low risk group, this represents an increase in deaths from 2.5 to 4.1 per 1000 detected cases. Conclusions The probability that the risk of mortality is increased by infection with VOC-202012/01 is high. If this finding is generalisable to other populations, infection with VOC-202012/1 has the potential to cause substantial additional mortality compared with previously circulating variants. Healthcare capacity planning and national and international control policies are all impacted by this finding, with increased mortality lending weight to the argument that further coordinated and stringent measures are justified to reduce deaths from SARS-CoV-2.

Abstract: The development and the present state of the “tps” series of software for use in geometric morphometrics on Windows-based computers are described. These programs have been used in hundreds of studies in mammals and other organisms. Download the complete issue.

Abstract: Response inhibition is essential for navigating everyday life. Its derailment is considered integral to numerous neurological and psychiatric disorders, and more generally, to a wide range of behavioral and health problems. Response-inhibition efficiency furthermore correlates with treatment outcome in some of these conditions. The stop-signal task is an essential tool to determine how quickly response inhibition is implemented. Despite its apparent simplicity, there are many features (ranging from task design to data analysis) that vary across studies in ways that can easily compromise the validity of the obtained results. Our goal is to facilitate a more accurate use of the stop-signal task. To this end, we provide 12 easy-to-implement consensus recommendations and point out the problems that can arise when they are not followed. Furthermore, we provide user-friendly open-source resources intended to inform statistical-power considerations, facilitate the correct implementation of the task, and assist in proper data analysis.

Abstract: BackgroundIn some patients with moderate-to-severe asthma, particularly those with noneosinophilic inflammation, the disease remains uncontrolled. This trial evaluated the efficacy and safety of tezepelumab (AMG 157/MEDI9929), a human monoclonal antibody specific for the epithelial-cell–derived cytokine thymic stromal lymphopoietin (TSLP), in patients whose asthma remained uncontrolled despite treatment with long-acting beta-agonists and medium-to-high doses of inhaled glucocorticoids. MethodsIn this phase 2, randomized, double-blind, placebo-controlled trial, we compared subcutaneous tezepelumab at three dose levels with placebo over a 52-week treatment period. The primary end point was the annualized rate of asthma exacerbations (events per patient-year) at week 52. ResultsThe use of tezepelumab at a dose of 70 mg every 4 weeks (low dose; 145 patients), 210 mg every 4 weeks (medium dose; 145 patients), or 280 mg every 2 weeks (high dose; 146 patients) resulted in annualized asthma exacerbation rates at ...

Abstract: Self-supervision provides effective representations for downstream tasks without requiring labels. However, existing approaches lag behind fully supervised training and are often not thought beneficial beyond obviating or reducing the need for annotations. We find that self-supervision can benefit robustness in a variety of ways, including robustness to adversarial examples, label corruption, and common input corruptions. Additionally, self-supervision greatly benefits out-of-distribution detection on difficult, near-distribution outliers, so much so that it exceeds the performance of fully supervised methods. These results demonstrate the promise of self-supervision for improving robustness and uncertainty estimation and establish these tasks as new axes of evaluation for future self-supervised learning research.

Abstract: Purpose Docetaxel added to androgen-deprivation therapy (ADT) significantly increases the longevity of some patients with metastatic hormone-sensitive prostate cancer. Herein, we present the outcomes of the CHAARTED (Chemohormonal Therapy Versus Androgen Ablation Randomized Trial for Extensive Disease in Prostate Cancer) trial with more mature follow-up and focus on tumor volume. Patients and Methods In this phase III study, 790 patients with metastatic hormone-sensitive prostate cancer were equally randomly assigned to receive either ADT in combination with docetaxel 75 mg/m2 for up to six cycles or ADT alone. The primary end point of the study was overall survival (OS). Additional analyses of the prospectively defined low- and high-volume disease subgroups were performed. High-volume disease was defined as presence of visceral metastases and/or ≥ four bone metastases with at least one outside of the vertebral column and pelvis. Results At a median follow-up of 53.7 months, the median OS was 57.6 months for the chemohormonal therapy arm versus 47.2 months for ADT alone (hazard ratio [HR], 0.72; 95% CI, 0.59 to 0.89; P = .0018). For patients with high-volume disease (n = 513), the median OS was 51.2 months with chemohormonal therapy versus 34.4 months with ADT alone (HR, 0.63; 95% CI, 0.50 to 0.79; P < .001). For those with low-volume disease (n = 277), no OS benefit was observed (HR, 1.04; 95% CI, 0.70 to 1.55; P = .86). Conclusion The clinical benefit from chemohormonal therapy in prolonging OS was confirmed for patients with high-volume disease; however, for patients with low-volume disease, no OS benefit was discerned.

Abstract: The volume of convolutional neural network (CNN) models proposed for face recognition has been continuously growing larger to better fit the large amount of training data. When training data are obtained from the Internet, the labels are likely to be ambiguous and inaccurate. This paper presents a Light CNN framework to learn a compact embedding on the large-scale face data with massive noisy labels. First, we introduce a variation of maxout activation, called max-feature-map (MFM), into each convolutional layer of CNN. Different from maxout activation that uses many feature maps to linearly approximate an arbitrary convex activation function, MFM does so via a competitive relationship. MFM can not only separate noisy and informative signals but also play the role of feature selection between two feature maps. Second, three networks are carefully designed to obtain better performance, meanwhile, reducing the number of parameters and computational costs. Finally, a semantic bootstrapping method is proposed to make the prediction of the networks more consistent with noisy labels. Experimental results show that the proposed framework can utilize large-scale noisy data to learn a Light model that is efficient in computational costs and storage spaces. The learned single network with a 256-D representation achieves state-of-the-art results on various face benchmarks without fine-tuning.