From the Cover: Ramification of stream networks

The geometric complexity of stream networks has been a source of fascination for centuries. However, a comprehensive understanding of ramification—the mechanism of branching by which such networks grow—remains elusive. Here we show that streams incised by groundwater seepage branch at a characteristic angle of 2π/5 = 72°. Our theory represents streams as a collection of paths growing and bifurcating in a diffusing field. Our observations of nearly 5,000 bifurcated streams growing in a 100 km² groundwater field on the Florida Panhandle yield a mean bifurcation angle of 71.9° ± 0.8°. This good accord between theory and observation suggests that the network geometry is determined by the external flow field but not, as classical theories imply, by the flow within the streams themselves.     

From the Cover: Fiber networks amplify active stress

Large-scale force generation is essential for biological functions such as cell motility, embryonic development, and muscle contraction. In these processes, forces generated at the molecular level by motor proteins are transmitted by disordered fiber networks, resulting in large-scale active stresses. Although these fiber networks are well characterized macroscopically, this stress generation by microscopic active units is not well understood. Here we theoretically study force transmission in these networks. We find that collective fiber buckling in the vicinity of a local active unit results in a rectification of stress towards strongly amplified isotropic contraction. This stress amplification is reinforced by the networks’ disordered nature, but saturates for high densities of active units. Our predictions are quantitatively consistent with experiments on reconstituted tissues and actomyosin networks and shed light on the role of the network microstructure in shaping active stresses in cells and tissue.Living systems constantly convert biochemical energy into forces and motion. In cells, forces are largely generated internally by molecular motors acting on the cytoskeleton, a scaffold of protein fibers (Fig. 1A). Forces from multiple motors are propagated along this fiber network, driving numerous processes such as mitosis and cell motility (1) and allowing the cell as a whole to exert stresses on its surroundings. At the larger scale of connective tissue, many such stress-exerting cells act on another type of fiber network known as the extracellular matrix (Fig. 1B). This network propagates cellular forces to the scale of the whole tissue, powering processes such as wound healing and morphogenesis. Despite important differences in molecular details and length scales, a common physical principle thus governs stress generation in biological matter: Internal forces from multiple localized “active units”—motors or cells—are propagated by a fiber network to generate large-scale stresses. However, a theoretical framework relating microscopic internal active forces to macroscopic stresses in these networks is lacking. Here we propose such a theory for elastic networks.Open in a separate windowFig. 1.Biological fiber networks (green) transmit forces generated by localized active units (red). (A) Myosin molecular motors exert forces on the actin cytoskeleton. (B) Contractile cells exert forces on the extracellular matrix. (C) The large nonlinear deformations induced by a model active unit in the surrounding fiber network result in stress amplification, as shown in this paper. Fiber color code is shown in D. (D) Each bond in the network comprises two rigid segments hinged together to allow buckling.This generic stress generation problem is confounded by the interplay of network disorder and nonlinear elasticity. Active units generate forces at the scale of the network mesh size, and force transmission to larger scales thus sensitively depends on local network heterogeneities. In the special case of linear elastic networks, the macroscopic active stress is simply given by the density of active force dipoles, irrespective of network characteristics (2). Importantly, however, this relationship is not applicable to most biological systems, because typical active forces are amply sufficient to probe the nonlinear properties of their constitutive fibers, which stiffen under tension and buckle under compression (3). Indeed, recent experiments on reconstituted biopolymer gels have shown that individual active units induce widespread buckling and stiffening (4, 5), and theory suggests that such fiber nonlinearities can enhance the range of force propagation (6, 7).Fiber networks also exhibit complex, nonlinear mechanical properties arising at larger scales, owing to collective deformations favored by the networks’ weak connectivity (3, 8). The role of connectivity in elasticity was famously investigated by Maxwell, who noticed that a spring network in dimension d becomes mechanically unstable for connectivities z < 2d. Interestingly, most biological fiber networks exhibit connectivities well below this threshold and therefore cannot be stabilized solely by the longitudinal stretching rigidity of their fibers. Instead, their macroscopic mechanical properties are typically controlled by the fiber bending rigidity (9). In contrast to stretching-dominated networks with connectivities above the Maxwell threshold, such weakly connected, bending-dominated networks are soft and extremely sensitive to mechanical perturbations (9–11). In these networks, stresses generated by active units propagate along intricate force chains (12, 13) whose effects on force transmission remain unexplored. Collections of such active units generate large stresses, with dramatic effects such as macroscopic network stiffening (14–16) and network remodeling (5, 17).Here we study the theoretical principles underlying stress generation by localized active units embedded in disordered fiber networks (Fig. 1C). We find that arbitrary local force distributions generically induce large isotropic, contractile stress fields at the network level, provided that the active forces are large enough to induce buckling in the network. In this case, the stress generated in a biopolymer network dramatically exceeds the stress level that would be produced in a linear elastic medium (2), implying a striking network-induced amplification of active stress. Our findings elucidate the origins and magnitude of stress amplification observed in experiments on reconstituted tissues (4, 18) and actomyosin networks (14, 17). We thus provide a conceptual framework for stress generation in biological fiber networks.     

From the Cover: Global view of the protein universe

To explore protein space from a global perspective, we consider 9,710 SCOP (Structural Classification of Proteins) domains with up to 70% sequence identity and present all similarities among them as networks: In the “domain network,” nodes represent domains, and edges connect domains that share “motifs,” i.e., significantly sized segments of similar sequence and structure. We explore the dependence of the network on the thresholds that define the evolutionary relatedness of the domains. At excessively strict thresholds the network falls apart completely; for very lax thresholds, there are network paths between virtually all domains. Interestingly, at intermediate thresholds the network constitutes two regions that can be described as “continuous” versus “discrete.” The continuous region comprises a large connected component, dominated by domains with alternating alpha and beta elements, and the discrete region includes the rest of the domains in isolated islands, each generally corresponding to a fold. We also construct the “motif network,” in which nodes represent recurring motifs, and edges connect motifs that appear in the same domain. This network also features a large and highly connected component of motifs that originate from domains with alternating alpha/beta elements (and some all-alpha domains), and smaller isolated islands. Indeed, the motif network suggests that nature reuses such motifs extensively. The networks suggest evolutionary paths between domains and give hints about protein evolution and the underlying biophysics. They provide natural means of organizing protein space, and could be useful for the development of strategies for protein search and design.How are proteins related to each other? Which physicochemical considerations affect protein evolution and how? A global view of the protein universe may shed light on these fundamental questions. It could also suggest new strategies for protein search and design (1–3). However, forming a global picture of the protein universe is difficult because we have to piece it together from the many local glimpses that our empirical data and computational tools provide. In other words, a global picture needs to portray the relationships among all proteins, yet we only have evidence of such relationships among several proteins, based on the similarity between their sequences, structures, and functions. The considerable size of the Protein Data Bank (4) also complicates this task.In particular, an intensely debated question is whether protein space is “discrete” or “continuous” (2, 3, 5–10). These terms are loosely defined. Discrete implies that the global picture consists of separate, island-like, structural entities. In the hierarchical protein domains Structural Classification of Proteins (SCOP) (11) these entities are termed “folds,” and in the CATH database (12) they are called “topologies.” Alternatively, “continuous” implies that the space between these entities is generally populated by cross-fold similarities (e.g., refs. 2, 5, 6, 9, 13–15). If such similarities are abundant, then one must account for them when organizing and searching proteins (5, 8, 16). In support of the abundance of such similarities is the remarkable success of structure prediction methods that piece together predictions of protein fragments or larger protein segments (e.g., ref. 17).There are different approaches to forming a global view of the protein universe (18). The most significant efforts are the ones embodied in the hierarchical classifications CATH and SCOP. However, a hierarchy implicitly assumes that there are isolated regions in protein space. An alternative approach is to study the protein universe via maps––where domains are represented by points in two or three dimensions, placed so that the distances between them depend on the dissimilarity between their corresponding domains (e.g., refs. 19–21). By coloring the points according to domain characteristics, one can visually identify global properties of the protein universe (19, 20). However, a map representation in low-dimensional Euclidean space implicitly suggests that similarity among domains is transitive (i.e., that similarity within the pairs AB and BC implies that AC is similar too); we know that this is often not the case (6). Finally, a third approach to study protein space is via similarity and cooccurrence networks. In similarity networks, nodes typically represent protein domains and edges connect similar domains. Several successful studies of protein space capitalize on such networks (22, 23). Cooccurrence networks of protein domains, in which nodes represent domains and edges connect cooccurring domains, were also studied to better understand protein evolution (24–26).Here, we study the global nature of the protein universe using domain and motif networks (Fig. 1). To construct these networks, we identify evolutionary relationships among a representative set of SCOP domains; we relate two domains if they share a significantly sized part (denoted motif) with similar structure and sequence. Our analysis reveals that protein space is both discrete and continuous: SCOP domains of the all-alpha, all-beta, and alpha + beta classes, in which alpha and beta elements do not mix, mostly populate the discrete parts, whereas alpha/beta domains, with alternating alpha and beta segments, mostly populate the continuous ones. We also find that recurring motifs are very abundant; the motifs from the all-alpha and alpha/beta domains are the more abundant, and the more gregarious ones.Open in a separate windowFig. 1.Constructing the domain and motif networks. (A) The aligned protein segments, marked in colors, are the motifs. (B) In the domain network, edges connect domains that share similar motifs (e.g., domain d1wjga_ and d1vlua_ that share the cyan motif). (C) In the motif network, edges connect cooccurring motifs (e.g., the orange and cyan motifs cooccur in the d1vlua_ domain).     

From the Cover: Evolution of microbial markets

Biological market theory has been used successfully to explain cooperative behavior in many animal species. Microbes also engage in cooperative behaviors, both with hosts and other microbes, that can be described in economic terms. However, a market approach is not traditionally used to analyze these interactions. Here, we extend the biological market framework to ask whether this theory is of use to evolutionary biologists studying microbes. We consider six economic strategies used by microbes to optimize their success in markets. We argue that an economic market framework is a useful tool to generate specific and interesting predictions about microbial interactions, including the evolution of partner discrimination, hoarding strategies, specialized versus diversified mutualistic services, and the role of spatial structures, such as flocks and consortia. There is untapped potential for studying the evolutionary dynamics of microbial systems. Market theory can help structure this potential by characterizing strategic investment of microbes across a diversity of conditions.     

From the Cover: Scaling laws predict global microbial diversity

Scaling laws underpin unifying theories of biodiversity and are among the most predictively powerful relationships in biology. However, scaling laws developed for plants and animals often go untested or fail to hold for microorganisms. As a result, it is unclear whether scaling laws of biodiversity will span evolutionarily distant domains of life that encompass all modes of metabolism and scales of abundance. Using a global-scale compilation of ∼35,000 sites and ∼5.6⋅10⁶ species, including the largest ever inventory of high-throughput molecular data and one of the largest compilations of plant and animal community data, we show similar rates of scaling in commonness and rarity across microorganisms and macroscopic plants and animals. We document a universal dominance scaling law that holds across 30 orders of magnitude, an unprecedented expanse that predicts the abundance of dominant ocean bacteria. In combining this scaling law with the lognormal model of biodiversity, we predict that Earth is home to upward of 1 trillion (10¹²) microbial species. Microbial biodiversity seems greater than ever anticipated yet predictable from the smallest to the largest microbiome.The understanding of microbial biodiversity has rapidly transformed over the past decade. High-throughput sequencing and bioinformatics have expanded the catalog of microbial taxa by orders of magnitude, whereas the unearthing of new phyla is reshaping the tree of life (1–3). At the same time, discoveries of novel forms of metabolism have provided insight into how microbes persist in virtually all aquatic, terrestrial, engineered, and host-associated ecosystems (4, 5). However, this period of discovery has uncovered few, if any, general rules for predicting microbial biodiversity at scales of abundance that characterize, for example, the ∼10¹⁴ cells of bacteria that inhabit a single human or the ∼10³⁰ cells of bacteria and archaea estimated to inhabit Earth (6, 7). Such findings would aid the estimation of global species richness and reveal whether theories of biodiversity hold across all scales of abundance and whether so-called law-like patterns of biodiversity span the tree of life.A primary goal of ecology and biodiversity theory is to predict diversity, commonness, and rarity across evolutionarily distant taxa and scales of space, time, and abundance (8–10). This goal can hardly be achieved without accounting for the most abundant, widespread, and metabolically, taxonomically, and functionally diverse organisms on Earth (i.e., microorganisms). However, tests of biodiversity theory rarely include both microbial and macrobial datasets. At the same time, the study of microbial ecology has yet to uncover quantitative relationships that predict diversity, commonness, and rarity at the scale of host microbiomes and beyond. These unexplored opportunities leave the understanding of biodiversity limited to the most conspicuous species of plants and animals. This lack of synthesis has also resulted in the independent study of two phenomena that likely represent a single universal pattern. Specifically, these phenomena are the highly uneven distributions of abundance that underpin biodiversity theory (11) and the universal pattern of microbial commonness and rarity known as the microbial “rare biosphere” (12).Scaling laws provide a promising path to the unified understanding and prediction of biodiversity. Also referred to as power laws, the forms of these relationships, y ∼ x^z, predict linear rates of change under logarithmic transformation [i.e., log(y) ∼ zlog(x)] and hence, proportional changes across orders of magnitude. Scaling laws reveal how physiological, ecological, and evolutionary constraints hold across genomes, cells, organisms, and communities of greatly varying size (13–15). Among the most widely known are the scaling of metabolic rate (B) with body size [M; B = B_oM^3/4 (13)] and the rate at which species richness (i.e., number of species; S) scale with area [A; S = cA^z (16)]. These scaling laws are predicted by powerful ecological theories, although evidence suggests that they fail for microorganisms (17–19). Beyond area and body size, there is an equally general constraint on biodiversity, that is, the number of individuals in an assemblage (N). Often referred to as total abundance, N can range from less than 10 individuals in a given area to the nearly 10³⁰ cells of bacteria and archaea on Earth (6, 7). This expanse outstrips the 22 orders of magnitude that separate the mass of a Prochlorococcus cell (3⋅1⁻¹⁶ kg) from a blue whale (1.9·10⁵ kg) and the 26 orders of magnitude that result from measuring Earth’s surface area at a spatial grain equivalent to bacteria (5.1⋅10²⁶ μm²).Here, we consider whether N may be one of the most powerful constraints on commonness and rarity and one of the most expansive variables across which aspects of biodiversity could scale. Although N imposes an obvious constraint on the number of species (i.e., S ≤ N), empirical and theoretical studies suggest that S scales with N at a rate of 0.25–0.5 (i.e., S ∼ N^z and 0.25 ≤ z ≤ 0.5) (20–22). Importantly, this relationship applies to samples from different systems and does not pertain to cumulative patterns (e.g., collector’s curves), which are based on resampling (20–22). Recent studies have also shown that N constrains universal patterns of commonness and rarity by imposing a numerical constraint on how abundance varies among species, across space, and through time (23, 24). Most notably, greater N leads to increasingly uneven distributions and greater rarity. Hence, we expect greater N to correspond to an increasingly uneven distribution among a greater number of species, an increasing portion of which should be rare. However, the strength of the relationships, whether they differ between microbes and macrobes, and whether they conform to scaling laws across orders of magnitude are virtually unknown.If aspects of diversity, commonness, and rarity scale with N, then local- to global-scale predictions of microbial biodiversity could be within reach. Likewise, if these relationships are similar for microbes and macrobes, then we may be closer to a unified understanding of biodiversity than previously thought. To answer these questions, we compiled the largest publicly available microbial and macrobial datasets to date. These data include 20,376 sites of bacterial, archaeal, and microscopic fungal communities and 14,862 sites of tree, bird, and mammal communities. We focused on taxonomic aspects of biodiversity, including species richness (S), similarity in abundance among species (evenness), concentration of N among relatively low-abundance species (rarity), and number of individuals belonging to the most abundant species (absolute dominance, N_max). We use the resulting relationships to predict N_max and S in large microbiomes and make empirically supported and theoretically underpinned estimates for the number of microbial species on Earth.     

From the Cover: An ancient,conserved gene regulatory network led to the rise of oral venom systems

Oral venom systems evolved multiple times in numerous vertebrates enabling the exploitation of unique predatory niches. Yet how and when they evolved remains poorly understood. Up to now, most research on venom evolution has focused strictly on the toxins. However, using toxins present in modern day animals to trace the origin of the venom system is difficult, since they tend to evolve rapidly, show complex patterns of expression, and were incorporated into the venom arsenal relatively recently. Here we focus on gene regulatory networks associated with the production of toxins in snakes, rather than the toxins themselves. We found that overall venom gland gene expression was surprisingly well conserved when compared to salivary glands of other amniotes. We characterized the “metavenom network,” a network of ∼3,000 nonsecreted housekeeping genes that are strongly coexpressed with the toxins, and are primarily involved in protein folding and modification. Conserved across amniotes, this network was coopted for venom evolution by exaptation of existing members and the recruitment of new toxin genes. For instance, starting from this common molecular foundation, Heloderma lizards, shrews, and solenodon, evolved venoms in parallel by overexpression of kallikreins, which were common in ancestral saliva and induce vasodilation when injected, causing circulatory shock. Derived venoms, such as those of snakes, incorporated novel toxins, though still rely on hypotension for prey immobilization. These similarities suggest repeated cooption of shared molecular machinery for the evolution of oral venom in mammals and reptiles, blurring the line between truly venomous animals and their ancestors.

Venoms are proteinaceous mixtures that can be traced and quantified to distinct genomic loci, providing a level of genetic tractability that is rare in other traits (1–4). This advantage of venom systems provides insights into processes of molecular evolution that are otherwise difficult to obtain. For example, studies in cnidarians showed that gene duplication is an effective way to increase protein dosage in tissues where different ecological roles can give rise to different patterns of gene expression (2, 5). Studies of venom in snakes have allowed comparisons of the relative importance of sequence evolution vs. gene expression evolution, as well as how a lack of genetic constraint enables diversity in complex traits (6, 7).Despite the wealth of knowledge venoms have provided about general evolutionary processes, the common molecular basis for the evolution of venom systems themselves is unknown. Even in snakes, which have perhaps the best studied venom systems, very little is known about the molecular architecture of these systems at their origin (8, 9). Using toxin families present in modern snakes to understand evolution at its origin is difficult because toxins evolve rapidly, both in terms of sequence and gene expression (10, 11). Toxins experience varying degrees of selection and drift, complicating interpretations of evolutionary models (12), and estimation of gene family evolution is often inconsistent, varying with which part of the gene (exon or intron) is used to construct the phylogeny (13). Most importantly, present-day toxins became a part of the venom over time; this diminishes their utility in trying to understand events that lead to the rise of venom systems in the nonvenomous ancestors of snakes (14, 15).A gene coexpression network aims to identify genes that interact with one another based on common expression profiles (16). Groups of coexpressed genes that have similar expression patterns across samples are identified using hierarchical clustering and are placed in gene “modules” (17). Constructing a network and comparing expression profiles of modules across taxa can identify key drivers of phenotypic change, as well as aid in identifying initial genetic targets of natural selection (18, 19). Comparative analysis using gene coexpression networks allows us to distinguish between ancient genetic modules representing core cellular processes, evolving modules that give rise to lineage-specific differences, and highly flexible modules that have evolved differently in different taxa (20). Gene coexpression networks are also widely used to construct gene regulatory networks (GRNs) owing to their reliability in capturing biologically relevant interactions between genes, as well as their high power in reproducing known protein–protein interactions (21, 22).Here we focus on gene coexpression networks involved in the production of snake venom, rather than the venom toxins themselves. Using a coexpression network we characterized the genes associated with venom production, which we term the “metavenom network,” and determine its biological role. We traced the origin of this network to the common ancestor of amniotes, which suggests that the venom system originated from a conserved gene regulatory network. The conserved nature of the metavenom network across amniotes suggests that oral venom systems started with a common gene regulatory foundation, and underwent lineage-specific changes to give rise to diverse venom systems in snakes, lizards, and even mammals.     

From the Cover: Highly specialized microbial diversity in hyper-arid polar desert

The McMurdo Dry Valleys in Antarctica are a cold hyperarid polar desert that present extreme challenges to life. Here, we report a culture-independent survey of multidomain microbial biodiversity in McKelvey Valley, a pristine example of the coldest desert on Earth. We demonstrate that life has adapted to form highly-specialized communities in distinct lithic niches occurring concomitantly within this terrain. Endoliths and chasmoliths in sandstone displayed greatest diversity, whereas soil was relatively depauperate and lacked a significant photoautotrophic component, apart from isolated islands of hypolithic cyanobacterial colonization on quartz rocks in soil contact. Communities supported previously unreported polar bacteria and fungi, but archaea were absent from all niches. Lithic community structure did not vary significantly on a landscape scale and stochastic moisture input due to snowmelt resulted in increases in colonization frequency without significantly affecting diversity. The findings show that biodiversity near the cold-arid limit for life is more complex than previously appreciated, but communities lack variability probably due to the high selective pressures of this extreme environment.     

From the Cover: Fine-tuning gene networks using simple sequence repeats

The parameters in a complex synthetic gene network must be extensively tuned before the network functions as designed. Here, we introduce a simple and general approach to rapidly tune gene networks in Escherichia coli using hypermutable simple sequence repeats embedded in the spacer region of the ribosome binding site. By varying repeat length, we generated expression libraries that incrementally and predictably sample gene expression levels over a 1,000-fold range. We demonstrate the utility of the approach by creating a bistable switch library that programmatically samples the expression space to balance the two states of the switch, and we illustrate the need for tuning by showing that the switch’s behavior is sensitive to host context. Further, we show that mutation rates of the repeats are controllable in vivo for stability or for targeted mutagenesis—suggesting a new approach to optimizing gene networks via directed evolution. This tuning methodology should accelerate the process of engineering functionally complex gene networks.     

From the Cover: Anatomy of funded research in science

Seeking research funding is an essential part of academic life. Funded projects are primarily collaborative in nature through internal and external partnerships, but what role does funding play in the formulation of these partnerships? Here, by examining over 43,000 scientific projects funded over the past three decades by one of the major government research agencies in the world, we characterize how the funding landscape has changed and its impacts on the underlying collaboration networks across different scales. We observed rising inequality in the distribution of funding and that its effect was most noticeable at the institutional level—the leading universities diversified their collaborations and increasingly became the knowledge brokers in the collaboration network. Furthermore, it emerged that these leading universities formed a rich club (i.e., a cohesive core through their close ties) and this reliance among them seemed to be a determining factor for their research success, with the elites in the core overattracting resources but also rewarding in terms of both research breadth and depth. Our results reveal how collaboration networks organize in response to external driving forces, which can have major ramifications on future research strategy and government policy.Higher education institutions are nationally assessed in a periodic manner across the globe [examples include the Research Excellence Framework (www.ref.ac.uk) in the United Kingdom, Excellenzinitiative (mediathek.dfg.de/thema/die-exzellenzinitiative/) in Germany, and Star Metrics (https://www.starmetrics.nih.gov/) in the United States], and tremendous effort has been put in place in maximizing research output, because assessment outcomes often have a direct financial impact on an institution’s revenue (1). Bibliometrics are commonly used for this kind of performance evaluations (2–7), and the volume of grant income is also generally seen as a good indicator of performance. Although many studies have examined the collaboration patterns originating from publication information (8–14), little is known about the characteristics of project collaborations supported by research funding, which is undoubtedly a type of research output in its own right, but also the origin of other research outputs.The volume of funding is often subject to direct and indirect constraints arising from internal research strategies and different levels of policy set out by the funding bodies and ultimately by the national government. This manifests into different emphases on both the research area and mode of collaboration, and potentially influences the way we form a project team. We have already seen examples of adaptive changes in our collaboration practices. For instance, research in the science and engineering sector is said to be increasingly interorganizational (15). In addition, there are different theories on the factors that may affect the establishment of a collaboration and how well a research team operates (13, 16). Elite universities were recognized as catalysts for facilitating large-scale multipartner research collaborations (15), and multidisciplinary collaborations were found to have higher potential to foster research outcomes (17). As a result, the setup of a project consortium for a grant application might require considerable strategic planning, because who and how we collaborate with can potentially affect the outcome of a bid, and we are yet to fully understand the underlying mechanics and dynamics.To shed light into the relations between funding landscapes and scientific collaborations, we here examine over 43,000 projects funded between 1985 and 2013 by the Engineering and Physical Sciences Research Council (EPSRC), the government body in the United Kingdom that provides funding to universities to undertake research in engineering and physical sciences, including mathematics, chemistry, materials science, energy, information and communications technology, and innovative manufacturing. For each year, we constructed two different types of collaboration networks in which the nodes are investigators and their affiliations, respectively, and an edge represents a funded project partnership between two nodes. We applied a network-based approach to analyze the local and global interlinkage in these networks; the former was performed by calculating the degree of brokerage (18–21) of individual nodes, which gauges the connectivity in the neighborhood of a node. As for the global level, we calculated the rich-club coefficient (22, 23) of the network and characterized the members of such core structure using a recently introduced profiling technique (24). In addition, we explored how these patterns evolved over time with the total funding in each year and how they correlated with research performance. Our results allow us to gain an insight into how changes in the funding landscape shaped the way we form research partnerships, providing a case study that is highly reflective of other countries in the European Union and possibly other developed countries worldwide.     

From the Cover: Predicting Social Security numbers from public data

Information about an individual''s place and date of birth can be exploited to predict his or her Social Security number (SSN). Using only publicly available information, we observed a correlation between individuals'' SSNs and their birth data and found that for younger cohorts the correlation allows statistical inference of private SSNs. The inferences are made possible by the public availability of the Social Security Administration''s Death Master File and the widespread accessibility of personal information from multiple sources, such as data brokers or profiles on social networking sites. Our results highlight the unexpected privacy consequences of the complex interactions among multiple data sources in modern information economies and quantify privacy risks associated with information revelation in public forums.     

From the Cover: Convolutional networks for fast,energy-efficient neuromorphic computing

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware’s underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.The human brain is capable of remarkable acts of perception while consuming very little energy. The dream of brain-inspired computing is to build machines that do the same, requiring high-accuracy algorithms and efficient hardware to run those algorithms. On the algorithm front, building on classic work on backpropagation (1), the neocognitron (2), and convolutional networks (3), deep learning has made great strides in achieving human-level performance on a wide range of recognition tasks (4). On the hardware front, building on foundational work on silicon neural systems (5), neuromorphic computing, using novel architectural primitives, has recently demonstrated hardware capable of running 1 million neurons and 256 million synapses for extremely low power (just 70 mW at real-time operation) (6). Bringing these approaches together holds the promise of a new generation of embedded, real-time systems, but first requires reconciling key differences in the structure and operation between contemporary algorithms and hardware. Here, we introduce and demonstrate an approach we call Eedn, energy-efficient deep neuromorphic networks, which creates convolutional networks whose connections, neurons, and weights have been adapted to run inference tasks on neuromorphic hardware.For structure, typical convolutional networks place no constraints on filter sizes, whereas neuromorphic systems can take advantage of blockwise connectivity that limits filter sizes, thereby saving energy because weights can now be stored in local on-chip memory within dedicated neural cores. Here, we present a convolutional network structure that naturally maps to the efficient connection primitives used in contemporary neuromorphic systems. We enforce this connectivity constraint by partitioning filters into multiple groups and yet maintain network integration by interspersing layers whose filter support region is able to cover incoming features from many groups by using a small topographic size (7).For operation, contemporary convolutional networks typically use high precision ( ≥ 32-bit) neurons and synapses to provide continuous derivatives and support small incremental changes to network state, both formally required for backpropagation-based gradient learning. In comparison, neuromorphic designs can use one-bit spikes to provide event-based computation and communication (consuming energy only when necessary) and can use low-precision synapses to colocate memory with computation (keeping data movement local and avoiding off-chip memory bottlenecks). Here, we demonstrate that by introducing two constraints into the learning rule—binary-valued neurons with approximate derivatives and trinary-valued ({−1,0,1}) synapses—it is possible to adapt backpropagation to create networks directly implementable using energy efficient neuromorphic dynamics. This approach draws inspiration from the spiking neurons and low-precision synapses of the brain (8) and builds on work showing that deep learning can create networks with constrained connectivity (9), low-precision synapses (10, 11), low-precision neurons (12–14), or both low-precision synapses and neurons (15, 16). For input data, we use a first layer to transform multivalued, multichannel input into binary channels using convolution filters that are learned via backpropagation (12, 16) and whose output can be sent on chip in the form of spikes. These binary channels, intuitively akin to independent components (17) learned with supervision, provide a parallel distributed representation to carry out high-fidelity computation without the need for high-precision representation.Critically, we demonstrate that bringing the above innovations together allows us to create networks that approach state-of-the-art accuracy performing inference on eight standard datasets, running on a neuromorphic chip at between 1,200 and 2,600 frames/s (FPS), using between 25 and 275 mW. We further explore how our approach scales by simulating multichip configurations. Ease-of-use is achieved using training tools built from existing, optimized deep learning frameworks (18), with learned parameters mapped to hardware using a high-level deployment language (19). Although we choose the IBM TrueNorth chip (6) for our example deployment platform, the essence of our constructions can apply to other emerging neuromorphic approaches (20–23) and may lead to new architectures that incorporate deep learning and efficient hardware primitives from the ground up.     

From the Cover: Bacillus subtilis chromosome organization oscillates between two distinct patterns

Bacterial chromosomes have been found to possess one of two distinct patterns of spatial organization. In the first, called “ori-ter” and exemplified by Caulobacter crescentus, the chromosome arms lie side-by-side, with the replication origin and terminus at opposite cell poles. In the second, observed in slow-growing Escherichia coli (“left-ori-right”), the two chromosome arms reside in separate cell halves, on either side of a centrally located origin. These two patterns, rotated 90° relative to each other, appear to result from different segregation mechanisms. Here, we show that the Bacillus subtilis chromosome alternates between them. For most of the cell cycle, newly replicated origins are maintained at opposite poles with chromosome arms adjacent to each other, in an ori-ter configuration. Shortly after replication initiation, the duplicated origins move as a unit to midcell and the two unreplicated arms resolve into opposite cell halves, generating a left-ori-right pattern. The origins are then actively segregated toward opposite poles, resetting the cycle. Our data suggest that the condensin complex and the parABS partitioning system are the principal driving forces underlying this oscillatory cycle. We propose that the distinct organization patterns observed for bacterial chromosomes reflect a common organization–segregation mechanism, and that simple modifications to it underlie the unique patterns observed in different species.Central to reproduction is the faithful segregation of replicated chromosomes to daughter cells. In eukaryotes, DNA replication, chromosome condensation, and sister chromatid segregation are separated into distinct steps in the cell cycle that are safeguarded by checkpoint pathways. In bacteria, these processes occur concurrently, posing unique challenges to genome integrity and inheritance (1, 2). In the absence of temporal control, bacteria take advantage of spatial organization to promote faithful and efficient chromosome segregation. The organization of the chromosome dictates where the chromosome is replicated, and the factors that organize and compact the newly replicated DNA play a central role in its segregation (1, 2).Studies in different bacteria have revealed strikingly distinct patterns of chromosome organization that appear to arise from different segregation mechanisms. In Caulobacter crescentus and Vibrio cholerae chromosome I, the origin and terminus are located at opposite cell poles, with the two replication arms between them, in a pattern referred to as “ori-ter” (3–5). After replication initiation, one of the sister origins is held in place and the other is actively translocated to the opposite cell pole, regenerating the ori-ter organization in both daughter cells (5–11). By contrast, in slow-growing Escherichia coli, the origin is located in the middle of the nucleoid and the two replication arms reside in opposite cell halves, in a “left-ori-right” pattern (12, 13). Replication initiates at midcell, and the duplicated origins segregate to the quarter positions followed by the left and right arms on either side, regenerating the left-ori-right pattern in the two daughter cells.Although chromosome organization was first analyzed in the Gram-positive bacterium Bacillus subtilis, our understanding of the replication–segregation cycle in this bacterium has remained elusive. In pioneering studies, it was shown that during spore formation, the replicated origins reside at opposite cell poles and the termini at midcell in an ori-ter ter-ori organization (14–18). A similar ori-ter pattern was observed during vegetative growth (15, 19). However, in separate studies, DNA replication was found to initiate at a midcell-localized origin (20, 21). How these disparate patterns fit into a coherent replication–segregation cycle has never been addressed and motivated this study. Our analysis has revealed that the B. subtilis chromosome follows an unexpected and previously unidentified choreography during vegetative growth in which the organization alternates between ori-ter and left-ori-right patterns. Our data further suggest that the highly conserved partitioning system (parABS) and the structural maintenance of chromosomes (SMC) condensin complex, in conjunction with replication initiation, function as the core components for this oscillating cycle. We propose that this cycle enhances the efficiency of DNA replication and sister chromosome segregation and provides a unifying model for the diverse patterns of chromosome organization observed in bacteria.     

From the Cover: Persistent ecological shifts in marine molluscan assemblages across the end-Cretaceous mass extinction

Contemporary biodiversity loss and population declines threaten to push the biosphere toward a tipping point with irreversible effects on ecosystem composition and function. As a potential example of a global-scale regime shift in the geological past, we assessed ecological changes across the end-Cretaceous mass extinction based on molluscan assemblages at four well-studied sites. By contrasting preextinction and postextinction rank abundance and numerical abundance in 19 molluscan modes of life—each defined as a unique combination of mobility level, feeding mode, and position relative to the substrate—we find distinct shifts in ecospace utilization, which significantly exceed predictions from null models. The magnitude of change in functional traits relative to normal temporal fluctuations at far-flung sites indicates that molluscan assemblages shifted to differently structured systems and faunal response was global. The strengths of temporal ecological shifts, however, are mostly within the range of preextinction site-to-site variability, demonstrating that local ecological turnover was similar to geographic variation over a broad latitudinal range. In conjunction with varied site-specific temporal patterns of individual modes of life, these spatial and temporal heterogeneities argue against a concerted phase shift of molluscan assemblages from one well-defined regime to another. At a broader ecological level, by contrast, congruent tendencies emerge and suggest deterministic processes. These patterns comprise the well-known increase of deposit-feeding mollusks in postextinction assemblages and increases in predators and predator-resistant modes of life, i.e., those characterized by elevated mobility and infaunal life habits.Recognizing nonlinear responses and tipping points in complex biological systems has raised concerns over the effects of global change, the extinction of species, and future global-scale shifts in the state of ecosystems (1, 2). Severe perturbations of the earth system, coupled with significant losses of biodiversity, have occurred a few times in earth history. These episodes of mass extinction provide the opportunity to study the ecological and evolutionary dynamics of the earth system when exposed to critical stress. Mass extinctions do not only devastate biodiversity—they also fundamentally restructure the variety of functions performed by the biota (3, 4). The analysis of ancient mass extinctions allows establishing the circumstances and the degree of system-level change in the geological past and thus could be informative of future changes in ecosystems due to anthropogenically driven biodiversity loss.The extinction event at the Cretaceous–Paleogene boundary (KPB, 66 million years before present) was the most recent mass extinction with an estimated 70% species loss (5). Similar to other mass extinctions it was associated with a profound disruption of the global carbon cycle (6). The ultimate trigger was probably the impact of an asteroid at Chicxulub in present-day Mexico (7), whereas Deccan Trap volcanism may have been an additional stressor (8). The most likely proximate killing mechanism was a crisis in primary productivity and global collapse of food webs owing to the suppression of photosynthesis (9–11). Other factors with devastating effects for marine ecosystems may have been metal poisoning (12), the acidification of oceanic surface waters (13), and short-lived global cooling (14).Major biotic changes associated with regime shifts can involve diversity loss, changes in biomass and trophic interactions, and the establishment of novel species assemblages (1). The contrast among different states in ecosystems is usually caused by a shift in dominance among organisms with different modes of life (15). Here we quantify the ecological change across the KPB in shallow marine benthic soft-bottom assemblages. These fossil assemblages are dominated by bivalve and gastropod mollusks, which not only have an excellent fossil record but also represent many different modes of life (MOLs), yielding unique insights into the ecological dynamics of the extinction and subsequent recovery. We compare molluscan ecospace occupation—as defined by the mobility, feeding mechanisms, and living positions of species (16)—in latest Cretaceous (Late Maastrichtian) preextinction times with that in earliest Paleogene (Danian) postextinction times.Previous studies showed an Early Danian increase in infaunal deposit feeders and mobility levels in some environments at some sites (17–21), whereas at other sites, either infaunal or epifaunal suspension feeders dominated and mobility levels and benthic tiering structure displayed no trends (21–25). Increased predation pressure after the KPB has been inferred from the radiation of predatory carnivores, in particular neogastopods (3, 18, 21, 26–28), elevated gastropod drilling frequencies (29, 30), and a trend toward deeper burrowing in bivalves (31), but predatory interactions did not increase universally (21, 25). One study demonstrated that Cretaceous–Paleogene spatial variation in functional group composition exceeded any changes through time (32), but otherwise this topic is largely unexplored. Some workers have questioned the presence of general ecological patterns (25), whereas others argued that ecological effects were habitat specific, with significant restructuring occurring in offshore assemblages and siliciclastic environments but not in shallow subtidal habitats and oligotrophic carbonate settings (20, 21).We studied mollusk-dominated siliciclastic shelf ecosystems at four well-studied sites before and after the KPB (SI Text S1). At each site, sedimentological evidence suggests that external environmental conditions were similar before and after the KPB (SI Text S1 and SI Text S2). The successions at three sites (Brazos River, Bajada del Jagüel, and Seymour Island) formed in a middle-to-outer shelf environment, whereas the fourth site (San Ramón) represents a tide-dominated delta. We estimated the ecological importance of each MOL in pre- and postextinction assemblages by its respective proportion based on counts of individuals. For this purpose, all Late Maastrichtian samples at a site were combined into a preextinction assemblage and all Danian samples into a postextinction assemblage. Specimen-level data were not available for the Maastrichtian of Seymour Island, and here the KPB comparison resides on the number of occurrences, i.e., counts of the presence of species of a particular MOL. In addition to comparing aggregate pre- and postextinction assemblages we used permutation tests and ordination techniques to compare the between-sample variation in ecological structure before and after the extinction event. Specifically, we tested whether these ecosystems experienced large, temporally abrupt, and persistent changes in ecological structure across the KPB. First, we examined whether postextinction assemblages constitute a fundamentally different assembly of functional groups. Differences to preextinction assemblages can be expected from previous work, but it is not clear whether they were beyond those of ordinary background fluctuations. Second, we explored whether any ecological disparity of postextinction assemblages reflects the ecospace occupation displayed by those taxa of preextinction assemblages that survived the extinction event. This scenario would suggest extinctions as the primary cause of ecological shifts. Third, we analyzed how consistently ecological patterns changed across sites. Congruence would suggest that the ecological systems responded deterministically to environmental change. Finally, we contrasted faunal shifts in time with the site-to-site variability to evaluate the spatiotemporal dimensions of ecological change.To test the significance of our results, we generated random species assemblages from the individual samples at each site for which we then calculated their ecological dissimilarity. By repeating this procedure many times we obtained a null distribution of ecological dissimilarity for “pre-” and “postextinction” assemblages (Materials and Methods). Comparison of the observed pattern with the prediction of the permuted null model allowed us to evaluate the significance of ecological change.