Abstract: The blind application of machine learning runs the risk of amplifying biases present in data. Such a danger is facing us with word embedding, a popular framework to represent text data as vectors which has been used in many machine learning and natural language processing tasks. We show that even word embeddings trained on Google News articles exhibit female/male gender stereotypes to a disturbing extent. This raises concerns because their widespread use, as we describe, often tends to amplify these biases. Geometrically, gender bias is first shown to be captured by a direction in the word embedding. Second, gender neutral words are shown to be linearly separable from gender definition words in the word embedding. Using these properties, we provide a methodology for modifying an embedding to remove gender stereotypes, such as the association between between the words receptionist and female, while maintaining desired associations such as between the words queen and female. We define metrics to quantify both direct and indirect gender biases in embeddings, and develop algorithms to "debias" the embedding. Using crowd-worker evaluation as well as standard benchmarks, we empirically demonstrate that our algorithms significantly reduce gender bias in embeddings while preserving the its useful properties such as the ability to cluster related concepts and to solve analogy tasks. The resulting embeddings can be used in applications without amplifying gender bias.

Abstract: Background Rabies is a notoriously underreported and neglected disease of low-income countries. This study aims to estimate the public health and economic burden of rabies circulating in domestic dog populations, globally and on a country-by-country basis, allowing an objective assessment of how much this preventable disease costs endemic countries.

Abstract: This paper examines whether there is a threshold above which …nancial development no longer has a positive eect on economic growth. We use dierent empirical approaches to show that there can indeed be "too much" …nance. In particular, our results suggest that …nance starts having a negative eect on output growth when credit to the private sector reaches 100% of GDP. We show that our results are consistent with the "vanishing eect" of …nancial development and that they are not driven by output volatility, banking crises, low institutional quality, or by dierences in bank regulation and supervision.

Abstract: Circular RNAs, a novel class of endogenous noncoding RNAs, are characterized by their covalently closed loop structures without a 5′ cap or a 3′ Poly A tail. Although the mechanisms of circular RNAs’ generation and function are not fully clear, recent research has shown that circular RNAs may function as potential molecular markers for disease diagnosis and treatment and play an important role in the initiation and progression of human diseases, especially in tumours. This review summarizes some information about categories, biogenesis, functions at the molecular level, properties of circular RNAs and the possibility of circular RNAs as biomarkers in cancers.

Abstract: PRECIS is a tool to help trialists make design decisions consistent with the intended purpose of their trial. This paper gives guidance on how to use an improved, validated version, PRECIS-2, which has been developed with the help of over 80 international trialists, clinicians, and policymakers. Keeping the original simple wheel format, PRECIS-2 has nine domains—eligibility criteria, recruitment, setting, organisation, flexibility (delivery), flexibility (adherence), follow-up, primary outcome, and primary analysis—scored from 1 (very explanatory) to 5 (very pragmatic) to facilitate domain discussion and consensus. It is hoped PRECIS-2 will be valuable in supporting the explicit matching of design decisions to how the trial results are intended to be used

Abstract: Heart failure (HF) is a rapidly growing public health issue with an estimated prevalence of >37.7 million individuals globally. HF is a shared chronic phase of cardiac functional impairment secondary to many aetiologies, and patients with HF experience numerous symptoms that affect their quality of life, including dyspnoea, fatigue, poor exercise tolerance, and fluid retention. Although the underlying causes of HF vary according to sex, age, ethnicity, comorbidities, and environment, the majority of cases remain preventable. HF is associated with increased morbidity and mortality, and confers a substantial burden to the health-care system. HF is a leading cause of hospitalization among adults and the elderly. In the USA, the total medical costs for patients with HF are expected to rise from US$20.9 billion in 2012 to $53.1 billion by 2030. Improvements in the medical management of risk factors and HF have stabilized the incidence of this disease in many countries. In this Review, we provide an overview of the latest epidemiological data on HF, and propose future directions for reducing the ever-increasing HF burden.

Abstract: New achievements in the realm of nanoscience and innovative techniques of nanomedicine have moved micro/nanoparticles (MNPs) to the point of becoming actually useful for practical applications in the near future. Various differences between the extracellular and intracellular environments of cancerous and normal cells and the particular characteristics of tumors such as physicochemical properties, neovasculature, elasticity, surface electrical charge, and pH have motivated the design and fabrication of inventive “smart” MNPs for stimulus-responsive controlled drug release. These novel MNPs can be tailored to be responsive to pH variations, redox potential, enzymatic activation, thermal gradients, magnetic fields, light, and ultrasound (US), or can even be responsive to dual or multi-combinations of different stimuli. This unparalleled capability has increased their importance as site-specific controlled drug delivery systems (DDSs) and has encouraged their rapid development in recent years. An in-depth understanding of the underlying mechanisms of these DDS approaches is expected to further contribute to this groundbreaking field of nanomedicine. Smart nanocarriers in the form of MNPs that can be triggered by internal or external stimulus are summarized and discussed in the present review, including pH-sensitive peptides and polymers, redox-responsive micelles and nanogels, thermo- or magnetic-responsive nanoparticles (NPs), mechanical- or electrical-responsive MNPs, light or ultrasound-sensitive particles, and multi-responsive MNPs including dual stimuli-sensitive nanosheets of graphene. This review highlights the recent advances of smart MNPs categorized according to their activation stimulus (physical, chemical, or biological) and looks forward to future pharmaceutical applications.

Abstract: Understanding the mechanisms controlling community diversity, functions, succession, and biogeography is a central, but poorly understood, topic in ecology, particularly in microbial ecology. Although stochastic processes are believed to play nonnegligible roles in shaping community structure, their importance relative to deterministic processes is hotly debated. The importance of ecological stochasticity in shaping microbial community structure is far less appreciated. Some of the main reasons for such heavy debates are the difficulty in defining stochasticity and the diverse methods used for delineating stochasticity. Here, we provide a critical review and synthesis of data from the most recent studies on stochastic community assembly in microbial ecology. We then describe both stochastic and deterministic components embedded in various ecological processes, including selection, dispersal, diversification, and drift. We also describe different approaches for inferring stochasticity from observational diversity patterns and highlight experimental approaches for delineating ecological stochasticity in microbial communities. In addition, we highlight research challenges, gaps, and future directions for microbial community assembly research.

Abstract: The use of flying platforms such as unmanned aerial vehicles (UAVs), popularly known as drones, is rapidly growing. In particular, with their inherent attributes such as mobility, flexibility, and adaptive altitude, UAVs admit several key potential applications in wireless systems. On the one hand, UAVs can be used as aerial base stations to enhance coverage, capacity, reliability, and energy efficiency of wireless networks. On the other hand, UAVs can operate as flying mobile terminals within a cellular network. Such cellular-connected UAVs can enable several applications ranging from real-time video streaming to item delivery. In this paper, a comprehensive tutorial on the potential benefits and applications of UAVs in wireless communications is presented. Moreover, the important challenges and the fundamental tradeoffs in UAV-enabled wireless networks are thoroughly investigated. In particular, the key UAV challenges such as three-dimensional deployment, performance analysis, channel modeling, and energy efficiency are explored along with representative results. Then, open problems and potential research directions pertaining to UAV communications are introduced. Finally, various analytical frameworks and mathematical tools such as optimization theory, machine learning, stochastic geometry, transport theory, and game theory are described. The use of such tools for addressing unique UAV problems is also presented. In a nutshell, this tutorial provides key guidelines on how to analyze, optimize, and design UAV-based wireless communication systems.

Abstract: CONTEXT The metabolic syndrome has been identified as a target for dietary therapies to reduce risk of cardiovascular disease; however, the role of diet in the etiology of the metabolic syndrome is poorly understood. OBJECTIVE To assess the effect of a Mediterranean-style diet on endothelial function and vascular inflammatory markers in patients with the metabolic syndrome. DESIGN, SETTING, AND PATIENTS Randomized, single-blind trial conducted from June 2001 to January 2004 at a university hospital in Italy among 180 patients (99 men and 81 women) with the metabolic syndrome, as defined by the Adult Treatment Panel III. INTERVENTIONS Patients in the intervention group (n = 90) were instructed to follow a Mediterranean-style diet and received detailed advice about how to increase daily consumption of whole grains, fruits, vegetables, nuts, and olive oil; patients in the control group (n = 90) followed a prudent diet (carbohydrates, 50%-60%; proteins, 15%-20%; total fat, <30%). MAIN OUTCOME MEASURES Nutrient intake; endothelial function score as a measure of blood pressure and platelet aggregation response to l-arginine; lipid and glucose parameters; insulin sensitivity; and circulating levels of high-sensitivity C-reactive protein (hs-CRP) and interleukins 6 (IL-6), 7 (IL-7), and 18 (IL-18). RESULTS After 2 years, patients following the Mediterranean-style diet consumed more foods rich in monounsaturated fat, polyunsaturated fat, and fiber and had a lower ratio of omega-6 to omega-3 fatty acids. Total fruit, vegetable, and nuts intake (274 g/d), whole grain intake (103 g/d), and olive oil consumption (8 g/d) were also significantly higher in the intervention group (P<.001). The level of physical activity increased in both groups by approximately 60%, without difference between groups (P =.22). Mean (SD) body weight decreased more in patients in the intervention group (-4.0 [1.1] kg) than in those in the control group (-1.2 [0.6] kg) (P<.001). Compared with patients consuming the control diet, patients consuming the intervention diet had significantly reduced serum concentrations of hs-CRP (P =.01), IL-6 (P =.04), IL-7 (P = 0.4), and IL-18 (P = 0.3), as well as decreased insulin resistance (P<.001). Endothelial function score improved in the intervention group (mean [SD] change, +1.9 [0.6]; P<.001) but remained stable in the control group (+0.2 [0.2]; P =.33). At 2 years of follow-up, 40 patients in the intervention group still had features of the metabolic syndrome, compared with 78 patients in the control group (P<.001). CONCLUSION A Mediterranean-style diet might be effective in reducing the prevalence of the metabolic syndrome and its associated cardiovascular risk.

Abstract: Background Long-term care facilities are high-risk settings for severe outcomes from outbreaks of Covid-19, owing to both the advanced age and frequent chronic underlying health conditions...

Abstract: Importance Parkinson disease is the most common form of parkinsonism, a group of neurological disorders with Parkinson disease–like movement problems such as rigidity, slowness, and tremor. More than 6 million individuals worldwide have Parkinson disease. Observations Diagnosis of Parkinson disease is based on history and examination. History can include prodromal features (eg, rapid eye movement sleep behavior disorder, hyposmia, constipation), characteristic movement difficulty (eg, tremor, stiffness, slowness), and psychological or cognitive problems (eg, cognitive decline, depression, anxiety). Examination typically demonstrates bradykinesia with tremor, rigidity, or both. Dopamine transporter single-photon emission computed tomography can improve the accuracy of diagnosis when the presence of parkinsonism is uncertain. Parkinson disease has multiple disease variants with different prognoses. Individuals with a diffuse malignant subtype (9%-16% of individuals with Parkinson disease) have prominent early motor and nonmotor symptoms, poor response to medication, and faster disease progression. Individuals with mild motor-predominant Parkinson disease (49%-53% of individuals with Parkinson disease) have mild symptoms, a good response to dopaminergic medications (eg, carbidopa-levodopa, dopamine agonists), and slower disease progression. Other individuals have an intermediate subtype. For all patients with Parkinson disease, treatment is symptomatic, focused on improvement in motor (eg, tremor, rigidity, bradykinesia) and nonmotor (eg, constipation, cognition, mood, sleep) signs and symptoms. No disease-modifying pharmacologic treatments are available. Dopamine-based therapies typically help initial motor symptoms. Nonmotor symptoms require nondopaminergic approaches (eg, selective serotonin reuptake inhibitors for psychiatric symptoms, cholinesterase inhibitors for cognition). Rehabilitative therapy and exercise complement pharmacologic treatments. Individuals experiencing complications, such as worsening symptoms and functional impairment when a medication dose wears off (“off periods”), medication-resistant tremor, and dyskinesias, benefit from advanced treatments such as therapy with levodopa-carbidopa enteral suspension or deep brain stimulation. Palliative care is part of Parkinson disease management. Conclusions and Relevance Parkinson disease is a heterogeneous disease with rapidly and slowly progressive forms. Treatment involves pharmacologic approaches (typically with levodopa preparations prescribed with or without other medications) and nonpharmacologic approaches (such as exercise and physical, occupational, and speech therapies). Approaches such as deep brain stimulation and treatment with levodopa-carbidopa enteral suspension can help individuals with medication-resistant tremor, worsening symptoms when the medication wears off, and dyskinesias.

Abstract: Cells may die from accidental cell death (ACD) or regulated cell death (RCD). ACD is a biologically uncontrolled process, whereas RCD involves tightly structured signaling cascades and molecularly defined effector mechanisms. A growing number of novel non-apoptotic forms of RCD have been identified and are increasingly being implicated in various human pathologies. Here, we critically review the current state of the art regarding non-apoptotic types of RCD, including necroptosis, pyroptosis, ferroptosis, entotic cell death, netotic cell death, parthanatos, lysosome-dependent cell death, autophagy-dependent cell death, alkaliptosis and oxeiptosis. The in-depth comprehension of each of these lethal subroutines and their intercellular consequences may uncover novel therapeutic targets for the avoidance of pathogenic cell loss.

Abstract: Recently, graph neural networks (GNNs) have revolutionized the field of graph representation learning through effectively learned node embeddings, and achieved state-of-the-art results in tasks such as node classification and link prediction. However, current GNN methods are inherently flat and do not learn hierarchical representations of graphs---a limitation that is especially problematic for the task of graph classification, where the goal is to predict the label associated with an entire graph. Here we propose DiffPool, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. DiffPool learns a differentiable soft cluster assignment for nodes at each layer of a deep GNN, mapping nodes to a set of clusters, which then form the coarsened input for the next GNN layer. Our experimental results show that combining existing GNN methods with DiffPool yields an average improvement of 5-10% accuracy on graph classification benchmarks, compared to all existing pooling approaches, achieving a new state-of-the-art on four out of five benchmark data sets.

Abstract: This article updates the guidance published in 2015 for authors submitting papers to British Journal of Pharmacology (Curtis et al., 2015) and is intended to provide the rubric for peer review. Thus, it is directed towards authors, reviewers and editors. Explanations for many of the requirements were outlined previously and are not restated here. The new guidelines are intended to replace those published previously. The guidelines have been simplified for ease of understanding by authors, to make it more straightforward for peer reviewers to check compliance and to facilitate the curation of the journal's efforts to improve standards.

Abstract: Large, labeled datasets have driven deep learning methods to achieve expert-level performance on a variety of medical imaging tasks. We present CheXpert, a large dataset that contains 224,316 chest radiographs of 65,240 patients. We design a labeler to automatically detect the presence of 14 observations in radiology reports, capturing uncertainties inherent in radiograph interpretation. We investigate different approaches to using the uncertainty labels for training convolutional neural networks that output the probability of these observations given the available frontal and lateral radiographs. On a validation set of 200 chest radiographic studies which were manually annotated by 3 board-certified radiologists, we find that different uncertainty approaches are useful for different pathologies. We then evaluate our best model on a test set composed of 500 chest radiographic studies annotated by a consensus of 5 board-certified radiologists, and compare the performance of our model to that of 3 additional radiologists in the detection of 5 selected pathologies. On Cardiomegaly, Edema, and Pleural Effusion, the model ROC and PR curves lie above all 3 radiologist operating points. We release the dataset to the public as a standard benchmark to evaluate performance of chest radiograph interpretation models.

Abstract: The rapid spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has had a large impact on global health, travel, and economy. Therefore, preventative and therapeutic measures are urgently needed. Here, we isolated monoclonal antibodies from three convalescent coronavirus disease 2019 (COVID-19) patients using a SARS-CoV-2 stabilized prefusion spike protein. These antibodies had low levels of somatic hypermutation and showed a strong enrichment in VH1-69, VH3-30-3, and VH1-24 gene usage. A subset of the antibodies was able to potently inhibit authentic SARS-CoV-2 infection at a concentration as low as 0.007 micrograms per milliliter. Competition and electron microscopy studies illustrate that the SARS-CoV-2 spike protein contains multiple distinct antigenic sites, including several receptor-binding domain (RBD) epitopes as well as non-RBD epitopes. In addition to providing guidance for vaccine design, the antibodies described here are promising candidates for COVID-19 treatment and prevention.

Abstract: In the last 5 years there have been a large number of new time series classification algorithms proposed in the literature. These algorithms have been evaluated on subsets of the 47 data sets in the University of California, Riverside time series classification archive. The archive has recently been expanded to 85 data sets, over half of which have been donated by researchers at the University of East Anglia. Aspects of previous evaluations have made comparisons between algorithms difficult. For example, several different programming languages have been used, experiments involved a single train/test split and some used normalised data whilst others did not. The relaunch of the archive provides a timely opportunity to thoroughly evaluate algorithms on a larger number of datasets. We have implemented 18 recently proposed algorithms in a common Java framework and compared them against two standard benchmark classifiers (and each other) by performing 100 resampling experiments on each of the 85 datasets. We use these results to test several hypotheses relating to whether the algorithms are significantly more accurate than the benchmarks and each other. Our results indicate that only nine of these algorithms are significantly more accurate than both benchmarks and that one classifier, the collective of transformation ensembles, is significantly more accurate than all of the others. All of our experiments and results are reproducible: we release all of our code, results and experimental details and we hope these experiments form the basis for more robust testing of new algorithms in the future.

Abstract: Despite the great advances made by deep learning in many machine learning problems, there is a relative dearth of deep learning approaches for anomaly detection. Those approaches which do exist involve networks trained to perform a task other than anomaly detection, namely generative models or compression, which are in turn adapted for use in anomaly detection; they are not trained on an anomaly detection based objective. In this paper we introduce a new anomaly detection method—Deep Support Vector Data Description—, which is trained on an anomaly detection based objective. The adaptation to the deep regime necessitates that our neural network and training procedure satisfy certain properties, which we demonstrate theoretically. We show the effectiveness of our method on MNIST and CIFAR-10 image benchmark datasets as well as on the detection of adversarial examples of GTSRB stop signs.

Abstract: In recent years, Discriminative Correlation Filter (DCF) based methods have significantly advanced the state-of-the-art in tracking. However, in the pursuit of ever increasing tracking performance, their characteristic speed and real-time capability have gradually faded. Further, the increasingly complex models, with massive number of trainable parameters, have introduced the risk of severe over-fitting. In this work, we tackle the key causes behind the problems of computational complexity and over-fitting, with the aim of simultaneously improving both speed and performance. We revisit the core DCF formulation and introduce: (i) a factorized convolution operator, which drastically reduces the number of parameters in the model; (ii) a compact generative model of the training sample distribution, that significantly reduces memory and time complexity, while providing better diversity of samples; (iii) a conservative model update strategy with improved robustness and reduced complexity. We perform comprehensive experiments on four benchmarks: VOT2016, UAV123, OTB-2015, and TempleColor. When using expensive deep features, our tracker provides a 20-fold speedup and achieves a 13.0% relative gain in Expected Average Overlap compared to the top ranked method in the VOT2016 challenge. Moreover, our fast variant, using hand-crafted features, operates at 60 Hz on a single CPU, while obtaining 65.0% AUC on OTB-2015.

Abstract: Tocilizumab (TCZ), a monoclonal antibody against interleukin-6 (IL-6), emerged as an alternative treatment for COVID-19 patients with a risk of cytokine storms recently. In the present study, we aimed to discuss the treatment response of TCZ therapy in COVID-19 infected patients. The demographic, treatment, laboratory parameters of C-reactive protein (CRP) and IL-6 before and after TCZ therapy and clinical outcome in the 15 COVID-19 patients were retrospectively assessed. Totally 15 patients with COVID-19 were included in this study. Two of them were moderately ill, six were seriously ill and seven were critically ill. The TCZ was used in combination with methylprednisolone in eight patients. Five patients received the TCZ administration twice or more. Although TCZ treatment ameliorated the increased CRP in all patients rapidly, for the four critically ill patients who received an only single dose of TCZ, three of them (No. 1, 2, and 3) still dead and the CRP level in the rest one patient (No. 7) failed to return to normal range with a clinical outcome of disease aggravation. Serum IL-6 level tended to further spiked firstly and then decreased after TCZ therapy in 10 patients. A persistent and dramatic increase of IL-6 was observed in these four patients who failed treatment. TCZ appears to be an effective treatment option in COVID-19 patients with a risk of cytokine storms. And for these critically ill patients with elevated IL-6, the repeated dose of the TCZ is recommended.

Abstract: The SARS-CoV-2 virus emerged in December 2019 and then spread rapidly worldwide, particularly to China, Japan, and South Korea. Scientists are endeavoring to find antivirals specific to the virus. Several drugs such as chloroquine, arbidol, remdesivir, and favipiravir are currently undergoing clinical studies to test their efficacy and safety in the treatment of coronavirus disease 2019 (COVID-19) in China; some promising results have been achieved thus far. This article summarizes agents with potential efficacy against SARS-CoV-2.

Abstract: We present ViLBERT (short for Vision-and-Language BERT), a model for learning task-agnostic joint representations of image content and natural language. We extend the popular BERT architecture to a multi-modal two-stream model, processing both visual and textual inputs in separate streams that interact through co-attentional transformer layers. We pretrain our model through two proxy tasks on the large, automatically collected Conceptual Captions dataset and then transfer it to multiple established vision-and-language tasks -- visual question answering, visual commonsense reasoning, referring expressions, and caption-based image retrieval -- by making only minor additions to the base architecture. We observe significant improvements across tasks compared to existing task-specific models -- achieving state-of-the-art on all four tasks. Our work represents a shift away from learning groundings between vision and language only as part of task training and towards treating visual grounding as a pretrainable and transferable capability.

Abstract: The growing commercial interest in indoor location-based services (ILBS) has spurred recent development of many indoor positioning techniques. Due to the absence of global positioning system (GPS) signal, many other signals have been proposed for indoor usage. Among them, Wi-Fi (802.11) emerges as a promising one due to the pervasive deployment of wireless LANs (WLANs). In particular, Wi-Fi fingerprinting has been attracting much attention recently because it does not require line-of-sight measurement of access points (APs) and achieves high applicability in complex indoor environment. This survey overviews recent advances on two major areas of Wi-Fi fingerprint localization: advanced localization techniques and efficient system deployment. Regarding advanced techniques to localize users, we present how to make use of temporal or spatial signal patterns, user collaboration, and motion sensors. Regarding efficient system deployment, we discuss recent advances on reducing offline labor-intensive survey, adapting to fingerprint changes, calibrating heterogeneous devices for signal collection, and achieving energy efficiency for smartphones. We study and compare the approaches through our deployment experiences, and discuss some future directions.

Abstract: Top-N sequential recommendation models each user as a sequence of items interacted in the past and aims to predict top-N ranked items that a user will likely interact in a »near future». The order of interaction implies that sequential patterns play an important role where more recent items in a sequence have a larger impact on the next item. In this paper, we propose a Convolutional Sequence Embedding Recommendation Model »Caser» as a solution to address this requirement. The idea is to embed a sequence of recent items into an »image» in the time and latent spaces and learn sequential patterns as local features of the image using convolutional filters. This approach provides a unified and flexible network structure for capturing both general preferences and sequential patterns. The experiments on public data sets demonstrated that Caser consistently outperforms state-of-the-art sequential recommendation methods on a variety of common evaluation metrics.

Abstract: Science and technologies based on terahertz frequency electromagnetic radiation (100 GHz–30 THz) have developed rapidly over the last 30 years. For most of the 20th Century, terahertz radiation, then referred to as sub-millimeter wave or far-infrared radiation, was mainly utilized by astronomers and some spectroscopists. Following the development of laser based terahertz time-domain spectroscopy in the 1980s and 1990s the field of THz science and technology expanded rapidly, to the extent that it now touches many areas from fundamental science to 'real world' applications. For example THz radiation is being used to optimize materials for new solar cells, and may also be a key technology for the next generation of airport security scanners. While the field was emerging it was possible to keep track of all new developments, however now the field has grown so much that it is increasingly difficult to follow the diverse range of new discoveries and applications that are appearing. At this point in time, when the field of THz science and technology is moving from an emerging to a more established and interdisciplinary field, it is apt to present a roadmap to help identify the breadth and future directions of the field. The aim of this roadmap is to present a snapshot of the present state of THz science and technology in 2017, and provide an opinion on the challenges and opportunities that the future holds. To be able to achieve this aim, we have invited a group of international experts to write 18 sections that cover most of the key areas of THz science and technology. We hope that The 2017 Roadmap on THz science and technology will prove to be a useful resource by providing a wide ranging introduction to the capabilities of THz radiation for those outside or just entering the field as well as providing perspective and breadth for those who are well established. We also feel that this review should serve as a useful guide for government and funding agencies.