@article{150, abstract = {A short, 14-amino-acid segment called SP1, located in the Gag structural protein1, has a critical role during the formation of the HIV-1 virus particle. During virus assembly, the SP1 peptide and seven preceding residues fold into a six-helix bundle, which holds together the Gag hexamer and facilitates the formation of a curved immature hexagonal lattice underneath the viral membrane2,3. Upon completion of assembly and budding, proteolytic cleavage of Gag leads to virus maturation, in which the immature lattice is broken down; the liberated CA domain of Gag then re-assembles into the mature conical capsid that encloses the viral genome and associated enzymes. Folding and proteolysis of the six-helix bundle are crucial rate-limiting steps of both Gag assembly and disassembly, and the six-helix bundle is an established target of HIV-1 inhibitors4,5. Here, using a combination of structural and functional analyses, we show that inositol hexakisphosphate (InsP6, also known as IP6) facilitates the formation of the six-helix bundle and assembly of the immature HIV-1 Gag lattice. IP6 makes ionic contacts with two rings of lysine residues at the centre of the Gag hexamer. Proteolytic cleavage then unmasks an alternative binding site, where IP6 interaction promotes the assembly of the mature capsid lattice. These studies identify IP6 as a naturally occurring small molecule that promotes both assembly and maturation of HIV-1.}, author = {Dick, Robert and Zadrozny, Kaneil K and Xu, Chaoyi and Schur, Florian and Lyddon, Terri D and Ricana, Clifton L and Wagner, Jonathan M and Perilla, Juan R and Ganser, Pornillos Barbie K and Johnson, Marc C and Pornillos, Owen and Vogt, Volker}, issn = {1476-4687}, journal = {Nature}, number = {7719}, pages = {509–512}, publisher = {Nature Publishing Group}, title = {{Inositol phosphates are assembly co-factors for HIV-1}}, doi = {10.1038/s41586-018-0396-4}, volume = {560}, year = {2018}, } @article{303, abstract = {The theory of tropical series, that we develop here, firstly appeared in the study of the growth of pluriharmonic functions. Motivated by waves in sandpile models we introduce a dynamic on the set of tropical series, and it is experimentally observed that this dynamic obeys a power law. So, this paper serves as a compilation of results we need for other articles and also introduces several objects interesting by themselves.}, author = {Kalinin, Nikita and Shkolnikov, Mikhail}, journal = {Discrete and Continuous Dynamical Systems- Series A}, number = {6}, pages = {2827 -- 2849}, publisher = {AIMS}, title = {{Introduction to tropical series and wave dynamic on them}}, doi = {10.3934/dcds.2018120}, volume = {38}, year = {2018}, } @inproceedings{14202, abstract = {Approximating a probability density in a tractable manner is a central task in Bayesian statistics. Variational Inference (VI) is a popular technique that achieves tractability by choosing a relatively simple variational family. Borrowing ideas from the classic boosting framework, recent approaches attempt to \emph{boost} VI by replacing the selection of a single density with a greedily constructed mixture of densities. In order to guarantee convergence, previous works impose stringent assumptions that require significant effort for practitioners. Specifically, they require a custom implementation of the greedy step (called the LMO) for every probabilistic model with respect to an unnatural variational family of truncated distributions. Our work fixes these issues with novel theoretical and algorithmic insights. On the theoretical side, we show that boosting VI satisfies a relaxed smoothness assumption which is sufficient for the convergence of the functional Frank-Wolfe (FW) algorithm. Furthermore, we rephrase the LMO problem and propose to maximize the Residual ELBO (RELBO) which replaces the standard ELBO optimization in VI. These theoretical enhancements allow for black box implementation of the boosting subroutine. Finally, we present a stopping criterion drawn from the duality gap in the classic FW analyses and exhaustive experiments to illustrate the usefulness of our theoretical and algorithmic contributions.}, author = {Locatello, Francesco and Dresdner, Gideon and Khanna, Rajiv and Valera, Isabel and Rätsch, Gunnar}, booktitle = {Advances in Neural Information Processing Systems}, isbn = {9781510884472}, issn = {1049-5258}, location = {Montreal, Canada}, publisher = {Neural Information Processing Systems Foundation}, title = {{Boosting black box variational inference}}, volume = {31}, year = {2018}, } @inproceedings{14201, abstract = {Variational inference is a popular technique to approximate a possibly intractable Bayesian posterior with a more tractable one. Recently, boosting variational inference has been proposed as a new paradigm to approximate the posterior by a mixture of densities by greedily adding components to the mixture. However, as is the case with many other variational inference algorithms, its theoretical properties have not been studied. In the present work, we study the convergence properties of this approach from a modern optimization viewpoint by establishing connections to the classic Frank-Wolfe algorithm. Our analyses yields novel theoretical insights regarding the sufficient conditions for convergence, explicit rates, and algorithmic simplifications. Since a lot of focus in previous works for variational inference has been on tractability, our work is especially important as a much needed attempt to bridge the gap between probabilistic models and their corresponding theoretical properties.}, author = {Locatello, Francesco and Khanna, Rajiv and Ghosh, Joydeep and Rätsch, Gunnar}, booktitle = {Proceedings of the 21st International Conference on Artificial Intelligence and Statistics}, location = {Playa Blanca, Lanzarote}, pages = {464--472}, publisher = {ML Research Press}, title = {{Boosting variational inference: An optimization perspective}}, volume = {84}, year = {2018}, } @inproceedings{14198, abstract = {High-dimensional time series are common in many domains. Since human cognition is not optimized to work well in high-dimensional spaces, these areas could benefit from interpretable low-dimensional representations. However, most representation learning algorithms for time series data are difficult to interpret. This is due to non-intuitive mappings from data features to salient properties of the representation and non-smoothness over time. To address this problem, we propose a new representation learning framework building on ideas from interpretable discrete dimensionality reduction and deep generative modeling. This framework allows us to learn discrete representations of time series, which give rise to smooth and interpretable embeddings with superior clustering performance. We introduce a new way to overcome the non-differentiability in discrete representation learning and present a gradient-based version of the traditional self-organizing map algorithm that is more performant than the original. Furthermore, to allow for a probabilistic interpretation of our method, we integrate a Markov model in the representation space. This model uncovers the temporal transition structure, improves clustering performance even further and provides additional explanatory insights as well as a natural representation of uncertainty. We evaluate our model in terms of clustering performance and interpretability on static (Fashion-)MNIST data, a time series of linearly interpolated (Fashion-)MNIST images, a chaotic Lorenz attractor system with two macro states, as well as on a challenging real world medical time series application on the eICU data set. Our learned representations compare favorably with competitor methods and facilitate downstream tasks on the real world data.}, author = {Fortuin, Vincent and Hüser, Matthias and Locatello, Francesco and Strathmann, Heiko and Rätsch, Gunnar}, booktitle = {International Conference on Learning Representations}, location = {New Orleans, LA, United States}, title = {{SOM-VAE: Interpretable discrete representation learning on time series}}, year = {2018}, } @inproceedings{14203, abstract = {We propose a conditional gradient framework for a composite convex minimization template with broad applications. Our approach combines smoothing and homotopy techniques under the CGM framework, and provably achieves the optimal O(1/k−−√) convergence rate. We demonstrate that the same rate holds if the linear subproblems are solved approximately with additive or multiplicative error. In contrast with the relevant work, we are able to characterize the convergence when the non-smooth term is an indicator function. Specific applications of our framework include the non-smooth minimization, semidefinite programming, and minimization with linear inclusion constraints over a compact domain. Numerical evidence demonstrates the benefits of our framework.}, author = {Yurtsever, Alp and Fercoq, Olivier and Locatello, Francesco and Cevher, Volkan}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, location = {Stockholm, Sweden}, pages = {5727--5736}, publisher = {ML Research Press}, title = {{A conditional gradient framework for composite convex minimization with applications to semidefinite programming}}, volume = {80}, year = {2018}, } @article{282, abstract = {Adaptive introgression is common in nature and can be driven by selection acting on multiple, linked genes. We explore the effects of polygenic selection on introgression under the infinitesimal model with linkage. This model assumes that the introgressing block has an effectively infinite number of genes, each with an infinitesimal effect on the trait under selection. The block is assumed to introgress under directional selection within a native population that is genetically homogeneous. We use individual-based simulations and a branching process approximation to compute various statistics of the introgressing block, and explore how these depend on parameters such as the map length and initial trait value associated with the introgressing block, the genetic variability along the block, and the strength of selection. Our results show that the introgression dynamics of a block under infinitesimal selection is qualitatively different from the dynamics of neutral introgression. We also find that in the long run, surviving descendant blocks are likely to have intermediate lengths, and clarify how the length is shaped by the interplay between linkage and infinitesimal selection. Our results suggest that it may be difficult to distinguish introgression of single loci from that of genomic blocks with multiple, tightly linked and weakly selected loci.}, author = {Sachdeva, Himani and Barton, Nicholas H}, journal = {Genetics}, number = {4}, pages = {1279 -- 1303}, publisher = {Genetics Society of America}, title = {{Introgression of a block of genome under infinitesimal selection}}, doi = {10.1534/genetics.118.301018}, volume = {209}, year = {2018}, } @inproceedings{108, abstract = {Universal hashing found a lot of applications in computer science. In cryptography the most important fact about universal families is the so called Leftover Hash Lemma, proved by Impagliazzo, Levin and Luby. In the language of modern cryptography it states that almost universal families are good extractors. In this work we provide a somewhat surprising characterization in the opposite direction. Namely, every extractor with sufficiently good parameters yields a universal family on a noticeable fraction of its inputs. Our proof technique is based on tools from extremal graph theory applied to the \'collision graph\' induced by the extractor, and may be of independent interest. We discuss possible applications to the theory of randomness extractors and non-malleable codes.}, author = {Obremski, Marciej and Skorski, Maciej}, location = {Vail, CO, USA}, publisher = {IEEE}, title = {{Inverted leftover hash lemma}}, doi = {10.1109/ISIT.2018.8437654}, volume = {2018}, year = {2018}, } @inproceedings{14204, abstract = {Two popular examples of first-order optimization methods over linear spaces are coordinate descent and matching pursuit algorithms, with their randomized variants. While the former targets the optimization by moving along coordinates, the latter considers a generalized notion of directions. Exploiting the connection between the two algorithms, we present a unified analysis of both, providing affine invariant sublinear O(1/t) rates on smooth objectives and linear convergence on strongly convex objectives. As a byproduct of our affine invariant analysis of matching pursuit, our rates for steepest coordinate descent are the tightest known. Furthermore, we show the first accelerated convergence rate O(1/t2) for matching pursuit and steepest coordinate descent on convex objectives.}, author = {Locatello, Francesco and Raj, Anant and Karimireddy, Sai Praneeth and Rätsch, Gunnar and Schölkopf, Bernhard and Stich, Sebastian U. and Jaggi, Martin}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {3198--3207}, publisher = {ML Research Press}, title = {{On matching pursuit and coordinate descent}}, volume = {80}, year = {2018}, } @inproceedings{160, abstract = {We present layered concurrent programs, a compact and expressive notation for specifying refinement proofs of concurrent programs. A layered concurrent program specifies a sequence of connected concurrent programs, from most concrete to most abstract, such that common parts of different programs are written exactly once. These programs are expressed in the ordinary syntax of imperative concurrent programs using gated atomic actions, sequencing, choice, and (recursive) procedure calls. Each concurrent program is automatically extracted from the layered program. We reduce refinement to the safety of a sequence of concurrent checker programs, one each to justify the connection between every two consecutive concurrent programs. These checker programs are also automatically extracted from the layered program. Layered concurrent programs have been implemented in the CIVL verifier which has been successfully used for the verification of several complex concurrent programs.}, author = {Kragl, Bernhard and Qadeer, Shaz}, location = {Oxford, UK}, pages = {79 -- 102}, publisher = {Springer}, title = {{Layered Concurrent Programs}}, doi = {10.1007/978-3-319-96145-3_5}, volume = {10981}, year = {2018}, } @article{280, abstract = {Flowers have a species-specific functional life span that determines the time window in which pollination, fertilization and seed set can occur. The stigma tissue plays a key role in flower receptivity by intercepting pollen and initiating pollen tube growth toward the ovary. In this article, we show that a developmentally controlled cell death programme terminates the functional life span of stigma cells in Arabidopsis. We identified the leaf senescence regulator ORESARA1 (also known as ANAC092) and the previously uncharacterized KIRA1 (also known as ANAC074) as partially redundant transcription factors that modulate stigma longevity by controlling the expression of programmed cell death-associated genes. KIRA1 expression is sufficient to induce cell death and terminate floral receptivity, whereas lack of both KIRA1 and ORESARA1 substantially increases stigma life span. Surprisingly, the extension of stigma longevity is accompanied by only a moderate extension of flower receptivity, suggesting that additional processes participate in the control of the flower's receptive life span.}, author = {Gao, Zhen and Daneva, Anna and Salanenka, Yuliya and Van Durme, Matthias and Huysmans, Marlies and Lin, Zongcheng and De Winter, Freya and Vanneste, Steffen and Karimi, Mansour and Van De Velde, Jan and Vandepoele, Klaas and Van De Walle, Davy and Dewettinck, Koen and Lambrecht, Bart and Nowack, Moritz}, journal = {Nature Plants}, number = {6}, pages = {365 -- 375}, publisher = {Nature Publishing Group}, title = {{KIRA1 and ORESARA1 terminate flower receptivity by promoting cell death in the stigma of Arabidopsis}}, doi = {10.1038/s41477-018-0160-7}, volume = {4}, year = {2018}, } @article{503, abstract = {Buffers are essential for diluting bacterial cultures for flow cytometry analysis in order to study bacterial physiology and gene expression parameters based on fluorescence signals. Using a variety of constitutively expressed fluorescent proteins in Escherichia coli K-12 strain MG1655, we found strong artifactual changes in fluorescence levels after dilution into the commonly used flow cytometry buffer phosphate-buffered saline (PBS) and two other buffer solutions, Tris-HCl and M9 salts. These changes appeared very rapidly after dilution, and were linked to increased membrane permeability and loss in cell viability. We observed buffer-related effects in several different E. coli strains, K-12, C and W, but not E. coli B, which can be partially explained by differences in lipopolysaccharide (LPS) and outer membrane composition. Supplementing the buffers with divalent cations responsible for outer membrane stability, Mg2+ and Ca2+, preserved fluorescence signals, membrane integrity and viability of E. coli. Thus, stabilizing the bacterial outer membrane is essential for precise and unbiased measurements of fluorescence parameters using flow cytometry.}, author = {Tomasek, Kathrin and Bergmiller, Tobias and Guet, Calin C}, journal = {Journal of Biotechnology}, pages = {40 -- 52}, publisher = {Elsevier}, title = {{Lack of cations in flow cytometry buffers affect fluorescence signals by reducing membrane stability and viability of Escherichia coli strains}}, doi = {10.1016/j.jbiotec.2018.01.008}, volume = {268}, year = {2018}, } @article{82, abstract = {In experimental cultures, when bacteria are mixed with lytic (virulent) bacteriophage, bacterial cells resistant to the phage commonly emerge and become the dominant population of bacteria. Following the ascent of resistant mutants, the densities of bacteria in these simple communities become limited by resources rather than the phage. Despite the evolution of resistant hosts, upon which the phage cannot replicate, the lytic phage population is most commonly maintained in an apparently stable state with the resistant bacteria. Several mechanisms have been put forward to account for this result. Here we report the results of population dynamic/evolution experiments with a virulent mutant of phage Lambda, λVIR, and Escherichia coli in serial transfer cultures. We show that, following the ascent of λVIR-resistant bacteria, λVIRis maintained in the majority of cases in maltose-limited minimal media and in all cases in nutrient-rich broth. Using mathematical models and experiments, we show that the dominant mechanism responsible for maintenance of λVIRin these resource-limited populations dominated by resistant E. coli is a high rate of either phenotypic or genetic transition from resistance to susceptibility—a hitherto undemonstrated mechanism we term "leaky resistance." We discuss the implications of leaky resistance to our understanding of the conditions for the maintenance of phage in populations of bacteria—their “existence conditions.”.}, author = {Chaudhry, Waqas and Pleska, Maros and Shah, Nilang and Weiss, Howard and Mccall, Ingrid and Meyer, Justin and Gupta, Animesh and Guet, Calin C and Levin, Bruce}, journal = {PLoS Biology}, number = {8}, publisher = {Public Library of Science}, title = {{Leaky resistance and the conditions for the existence of lytic bacteriophage}}, doi = {10.1371/journal.pbio.2005971}, volume = {16}, year = {2018}, } @article{4, abstract = {We present a data-driven technique to instantly predict how fluid flows around various three-dimensional objects. Such simulation is useful for computational fabrication and engineering, but is usually computationally expensive since it requires solving the Navier-Stokes equation for many time steps. To accelerate the process, we propose a machine learning framework which predicts aerodynamic forces and velocity and pressure fields given a threedimensional shape input. Handling detailed free-form three-dimensional shapes in a data-driven framework is challenging because machine learning approaches usually require a consistent parametrization of input and output. We present a novel PolyCube maps-based parametrization that can be computed for three-dimensional shapes at interactive rates. This allows us to efficiently learn the nonlinear response of the flow using a Gaussian process regression. We demonstrate the effectiveness of our approach for the interactive design and optimization of a car body.}, author = {Umetani, Nobuyuki and Bickel, Bernd}, journal = {ACM Trans. Graph.}, number = {4}, publisher = {ACM}, title = {{Learning three-dimensional flow for interactive aerodynamic design}}, doi = {10.1145/3197517.3201325}, volume = {37}, year = {2018}, } @inproceedings{183, abstract = {Fault-localization is considered to be a very tedious and time-consuming activity in the design of complex Cyber-Physical Systems (CPS). This laborious task essentially requires expert knowledge of the system in order to discover the cause of the fault. In this context, we propose a new procedure that AIDS designers in debugging Simulink/Stateflow hybrid system models, guided by Signal Temporal Logic (STL) specifications. The proposed method relies on three main ingredients: (1) a monitoring and a trace diagnostics procedure that checks whether a tested behavior satisfies or violates an STL specification, localizes time segments and interfaces variables contributing to the property violations; (2) a slicing procedure that maps these observable behavior segments to the internal states and transitions of the Simulink model; and (3) a spectrum-based fault-localization method that combines the previous analysis from multiple tests to identify the internal states and/or transitions that are the most likely to explain the fault. We demonstrate the applicability of our approach on two Simulink models from the automotive and the avionics domain.}, author = {Bartocci, Ezio and Ferrere, Thomas and Manjunath, Niveditha and Nickovic, Dejan}, location = {Porto, Portugal}, pages = {197 -- 206}, publisher = {Association for Computing Machinery, Inc}, title = {{Localizing faults in simulink/stateflow models with STL}}, doi = {10.1145/3178126.3178131}, year = {2018}, } @article{566, abstract = {We consider large random matrices X with centered, independent entries which have comparable but not necessarily identical variances. Girko's circular law asserts that the spectrum is supported in a disk and in case of identical variances, the limiting density is uniform. In this special case, the local circular law by Bourgade et. al. [11,12] shows that the empirical density converges even locally on scales slightly above the typical eigenvalue spacing. In the general case, the limiting density is typically inhomogeneous and it is obtained via solving a system of deterministic equations. Our main result is the local inhomogeneous circular law in the bulk spectrum on the optimal scale for a general variance profile of the entries of X. }, author = {Alt, Johannes and Erdös, László and Krüger, Torben H}, journal = {Annals Applied Probability }, number = {1}, pages = {148--203}, publisher = {Institute of Mathematical Statistics}, title = {{Local inhomogeneous circular law}}, doi = {10.1214/17-AAP1302}, volume = {28}, year = {2018}, } @article{106, abstract = {The goal of this article is to introduce the reader to the theory of intrinsic geometry of convex surfaces. We illustrate the power of the tools by proving a theorem on convex surfaces containing an arbitrarily long closed simple geodesic. Let us remind ourselves that a curve in a surface is called geodesic if every sufficiently short arc of the curve is length minimizing; if, in addition, it has no self-intersections, we call it simple geodesic. A tetrahedron with equal opposite edges is called isosceles. The axiomatic method of Alexandrov geometry allows us to work with the metrics of convex surfaces directly, without approximating it first by a smooth or polyhedral metric. Such approximations destroy the closed geodesics on the surface; therefore it is difficult (if at all possible) to apply approximations in the proof of our theorem. On the other hand, a proof in the smooth or polyhedral case usually admits a translation into Alexandrov’s language; such translation makes the result more general. In fact, our proof resembles a translation of the proof given by Protasov. Note that the main theorem implies in particular that a smooth convex surface does not have arbitrarily long simple closed geodesics. However we do not know a proof of this corollary that is essentially simpler than the one presented below.}, author = {Akopyan, Arseniy and Petrunin, Anton}, journal = {Mathematical Intelligencer}, number = {3}, pages = {26 -- 31}, publisher = {Springer}, title = {{Long geodesics on convex surfaces}}, doi = {10.1007/s00283-018-9795-5}, volume = {40}, year = {2018}, } @misc{9810, author = {Chaudhry, Waqas and Pleska, Maros and Shah, Nilang and Weiss, Howard and Mccall, Ingrid and Meyer, Justin and Gupta, Animesh and Guet, Calin C and Levin, Bruce}, publisher = {Public Library of Science}, title = {{Numerical data used in figures}}, doi = {10.1371/journal.pbio.2005971.s008}, year = {2018}, } @article{275, abstract = {Lymphatic endothelial cells (LECs) release extracellular chemokines to guide the migration of dendritic cells. In this study, we report that LECs also release basolateral exosome-rich endothelial vesicles (EEVs) that are secreted in greater numbers in the presence of inflammatory cytokines and accumulate in the perivascular stroma of small lymphatic vessels in human chronic inflammatory diseases. Proteomic analyses of EEV fractions identified > 1,700 cargo proteins and revealed a dominant motility-promoting protein signature. In vitro and ex vivo EEV fractions augmented cellular protrusion formation in a CX3CL1/fractalkine-dependent fashion and enhanced the directional migratory response of human dendritic cells along guidance cues. We conclude that perilymphatic LEC exosomes enhance exploratory behavior and thus promote directional migration of CX3CR1-expressing cells in complex tissue environments.}, author = {Brown, Markus and Johnson, Louise and Leone, Dario and Májek, Peter and Vaahtomeri, Kari and Senfter, Daniel and Bukosza, Nora and Schachner, Helga and Asfour, Gabriele and Langer, Brigitte and Hauschild, Robert and Parapatics, Katja and Hong, Young and Bennett, Keiryn and Kain, Renate and Detmar, Michael and Sixt, Michael K and Jackson, David and Kerjaschki, Dontscho}, journal = {Journal of Cell Biology}, number = {6}, pages = {2205 -- 2221}, publisher = {Rockefeller University Press}, title = {{Lymphatic exosomes promote dendritic cell migration along guidance cues}}, doi = {10.1083/jcb.201612051}, volume = {217}, year = {2018}, } @article{158, abstract = {The angiosperm seed is composed of three genetically distinct tissues: the diploid embryo that originates from the fertilized egg cell, the triploid endosperm that is produced from the fertilized central cell, and the maternal sporophytic integuments that develop into the seed coat1. At the onset of embryo development in Arabidopsis thaliana, the zygote divides asymmetrically, producing a small apical embryonic cell and a larger basal cell that connects the embryo to the maternal tissue2. The coordinated and synchronous development of the embryo and the surrounding integuments, and the alignment of their growth axes, suggest communication between maternal tissues and the embryo. In contrast to animals, however, where a network of maternal factors that direct embryo patterning have been identified3,4, only a few maternal mutations have been described to affect embryo development in plants5–7. Early embryo patterning in Arabidopsis requires accumulation of the phytohormone auxin in the apical cell by directed transport from the suspensor8–10. However, the origin of this auxin has remained obscure. Here we investigate the source of auxin for early embryogenesis and provide evidence that the mother plant coordinates seed development by supplying auxin to the early embryo from the integuments of the ovule. We show that auxin response increases in ovules after fertilization, due to upregulated auxin biosynthesis in the integuments, and this maternally produced auxin is required for correct embryo development.}, author = {Robert, Hélène and Park, Chulmin and Gutièrrez, Carla and Wójcikowska, Barbara and Pěnčík, Aleš and Novák, Ondřej and Chen, Junyi and Grunewald, Wim and Dresselhaus, Thomas and Friml, Jirí and Laux, Thomas}, journal = {Nature Plants}, number = {8}, pages = {548 -- 553}, publisher = {Nature Publishing Group}, title = {{Maternal auxin supply contributes to early embryo patterning in Arabidopsis}}, doi = {10.1038/s41477-018-0204-z}, volume = {4}, year = {2018}, }