@article{10270, abstract = {Plants develop new organs to adjust their bodies to dynamic changes in the environment. How independent organs achieve anisotropic shapes and polarities is poorly understood. To address this question, we constructed a mechano-biochemical model for Arabidopsis root meristem growth that integrates biologically plausible principles. Computer model simulations demonstrate how differential growth of neighboring tissues results in the initial symmetry-breaking leading to anisotropic root growth. Furthermore, the root growth feeds back on a polar transport network of the growth regulator auxin. Model, predictions are in close agreement with in vivo patterns of anisotropic growth, auxin distribution, and cell polarity, as well as several root phenotypes caused by chemical, mechanical, or genetic perturbations. Our study demonstrates that the combination of tissue mechanics and polar auxin transport organizes anisotropic root growth and cell polarities during organ outgrowth. Therefore, a mobile auxin signal transported through immobile cells drives polarity and growth mechanics to coordinate complex organ development.}, author = {Marconi, Marco and Gallemi, Marçal and Benková, Eva and Wabnik, Krzysztof}, issn = {2050-084X}, journal = {eLife}, publisher = {eLife Sciences Publications}, title = {{A coupled mechano-biochemical model for cell polarity guided anisotropic root growth}}, doi = {10.7554/elife.72132}, volume = {10}, year = {2021}, } @article{10280, abstract = {Machines enabled the Industrial Revolution and are central to modern technological progress: A machine’s parts transmit forces, motion, and energy to one another in a predetermined manner. Today’s engineering frontier, building artificial micromachines that emulate the biological machinery of living organisms, requires faithful assembly and energy consumption at the microscale. Here, we demonstrate the programmable assembly of active particles into autonomous metamachines using optical templates. Metamachines, or machines made of machines, are stable, mobile and autonomous architectures, whose dynamics stems from the geometry. We use the interplay between anisotropic force generation of the active colloids with the control of their orientation by local geometry. This allows autonomous reprogramming of active particles of the metamachines to achieve multiple functions. It permits the modular assembly of metamachines by fusion, reconfiguration of metamachines and, we anticipate, a shift in focus of self-assembly towards active matter and reprogrammable materials.}, author = {Aubret, Antoine and Martinet, Quentin and Palacci, Jérémie A}, issn = {2041-1723}, journal = {Nature Communications}, number = {1}, publisher = {Springer Nature}, title = {{Metamachines of pluripotent colloids}}, doi = {10.1038/s41467-021-26699-6}, volume = {12}, year = {2021}, } @article{10283, abstract = {During the past decade, the scientific community and outside observers have noted a concerning lack of rigor and transparency in preclinical research that led to talk of a “reproducibility crisis” in the life sciences (Baker, 2016; Bespalov & Steckler, 2018; Heddleston et al, 2021). Various measures have been proposed to address the problem: from better training of scientists to more oversight to expanded publishing practices such as preregistration of studies. The recently published EQIPD (Enhancing Quality in Preclinical Data) System is, to date, the largest initiative that aims to establish a systematic approach for increasing the robustness and reliability of biomedical research (Bespalov et al, 2021). However, promoting a cultural change in research practices warrants a broad adoption of the Quality System and its underlying philosophy. It is here that academic Core Facilities (CF), research service providers at universities and research institutions, can make a difference. It is fair to assume that a significant fraction of published data originated from experiments that were designed, run, or analyzed in CFs. These academic services play an important role in the research ecosystem by offering access to cutting-edge equipment and by developing and testing novel techniques and methods that impact research in the academic and private sectors alike (Bikovski et al, 2020). Equipment and infrastructure are not the only value: CFs employ competent personnel with profound knowledge and practical experience of the specific field of interest: animal behavior, imaging, crystallography, genomics, and so on. Thus, CFs are optimally positioned to address concerns about the quality and robustness of preclinical research.}, author = {Restivo, Leonardo and Gerlach, Björn and Tsoory, Michael and Bikovski, Lior and Badurek, Sylvia and Pitzer, Claudia and Kos-Braun, Isabelle C. and Mausset-Bonnefont, Anne Laure Mj and Ward, Jonathan and Schunn, Michael and Noldus, Lucas P.J.J. and Bespalov, Anton and Voikar, Vootele}, issn = {1469-3178}, journal = {EMBO Reports}, publisher = {EMBO Press}, title = {{Towards best practices in research: Role of academic core facilities}}, doi = {10.15252/embr.202153824}, volume = {22}, year = {2021}, } @article{10285, abstract = {We study the overlaps between right and left eigenvectors for random matrices of the spherical ensemble, as well as truncated unitary ensembles in the regime where half of the matrix at least is truncated. These two integrable models exhibit a form of duality, and the essential steps of our investigation can therefore be performed in parallel. In every case, conditionally on all eigenvalues, diagonal overlaps are shown to be distributed as a product of independent random variables with explicit distributions. This enables us to prove that the scaled diagonal overlaps, conditionally on one eigenvalue, converge in distribution to a heavy-tail limit, namely, the inverse of a γ2 distribution. We also provide formulae for the conditional expectation of diagonal and off-diagonal overlaps, either with respect to one eigenvalue, or with respect to the whole spectrum. These results, analogous to what is known for the complex Ginibre ensemble, can be obtained in these cases thanks to integration techniques inspired from a previous work by Forrester & Krishnapur.}, author = {Dubach, Guillaume}, issn = {1083-6489}, journal = {Electronic Journal of Probability}, publisher = {Institute of Mathematical Statistics}, title = {{On eigenvector statistics in the spherical and truncated unitary ensembles}}, doi = {10.1214/21-EJP686}, volume = {26}, year = {2021}, } @article{10290, abstract = {A precise quantitative description of the ultrastructural characteristics underlying biological mechanisms is often key to their understanding. This is particularly true for dynamic extra- and intracellular filamentous assemblies, playing a role in cell motility, cell integrity, cytokinesis, tissue formation and maintenance. For example, genetic manipulation or modulation of actin regulatory proteins frequently manifests in changes of the morphology, dynamics, and ultrastructural architecture of actin filament-rich cell peripheral structures, such as lamellipodia or filopodia. However, the observed ultrastructural effects often remain subtle and require sufficiently large datasets for appropriate quantitative analysis. The acquisition of such large datasets has been enabled by recent advances in high-throughput cryo-electron tomography (cryo-ET) methods. This also necessitates the development of complementary approaches to maximize the extraction of relevant biological information. We have developed a computational toolbox for the semi-automatic quantification of segmented and vectorized filamentous networks from pre-processed cryo-electron tomograms, facilitating the analysis and cross-comparison of multiple experimental conditions. GUI-based components simplify the processing of data and allow users to obtain a large number of ultrastructural parameters describing filamentous assemblies. We demonstrate the feasibility of this workflow by analyzing cryo-ET data of untreated and chemically perturbed branched actin filament networks and that of parallel actin filament arrays. In principle, the computational toolbox presented here is applicable for data analysis comprising any type of filaments in regular (i.e. parallel) or random arrangement. We show that it can ease the identification of key differences between experimental groups and facilitate the in-depth analysis of ultrastructural data in a time-efficient manner.}, author = {Dimchev, Georgi A and Amiri, Behnam and Fäßler, Florian and Falcke, Martin and Schur, Florian KM}, issn = {1047-8477}, journal = {Journal of Structural Biology}, keywords = {Structural Biology}, number = {4}, publisher = {Elsevier }, title = {{Computational toolbox for ultrastructural quantitative analysis of filament networks in cryo-ET data}}, doi = {10.1016/j.jsb.2021.107808}, volume = {213}, year = {2021}, } @article{10301, abstract = {De novo protein synthesis is required for synapse modifications underlying stable memory encoding. Yet neurons are highly compartmentalized cells and how protein synthesis can be regulated at the synapse level is unknown. Here, we characterize neuronal signaling complexes formed by the postsynaptic scaffold GIT1, the mechanistic target of rapamycin (mTOR) kinase, and Raptor that couple synaptic stimuli to mTOR-dependent protein synthesis; and identify NMDA receptors containing GluN3A subunits as key negative regulators of GIT1 binding to mTOR. Disruption of GIT1/mTOR complexes by enhancing GluN3A expression or silencing GIT1 inhibits synaptic mTOR activation and restricts the mTOR-dependent translation of specific activity-regulated mRNAs. Conversely, GluN3A removal enables complex formation, potentiates mTOR-dependent protein synthesis, and facilitates the consolidation of associative and spatial memories in mice. The memory enhancement becomes evident with light or spaced training, can be achieved by selectively deleting GluN3A from excitatory neurons during adulthood, and does not compromise other aspects of cognition such as memory flexibility or extinction. Our findings provide mechanistic insight into synaptic translational control and reveal a potentially selective target for cognitive enhancement.}, author = {Conde-Dusman, María J and Dey, Partha N and Elía-Zudaire, Óscar and Garcia Rabaneda, Luis E and García-Lira, Carmen and Grand, Teddy and Briz, Victor and Velasco, Eric R and Andero Galí, Raül and Niñerola, Sergio and Barco, Angel and Paoletti, Pierre and Wesseling, John F and Gardoni, Fabrizio and Tavalin, Steven J and Perez-Otaño, Isabel}, issn = {2050-084X}, journal = {eLife}, keywords = {general immunology and microbiology, general biochemistry, genetics and molecular biology, general medicine, general neuroscience}, publisher = {eLife Sciences Publications}, title = {{Control of protein synthesis and memory by GluN3A-NMDA receptors through inhibition of GIT1/mTORC1 assembly}}, doi = {10.7554/elife.71575}, volume = {10}, year = {2021}, } @article{10310, abstract = {A high-resolution structure of trimeric cyanobacterial Photosystem I (PSI) from Thermosynechococcus elongatus was reported as the first atomic model of PSI almost 20 years ago. However, the monomeric PSI structure has not yet been reported despite long-standing interest in its structure and extensive spectroscopic characterization of the loss of red chlorophylls upon monomerization. Here, we describe the structure of monomeric PSI from Thermosynechococcus elongatus BP-1. Comparison with the trimer structure gave detailed insights into monomerization-induced changes in both the central trimerization domain and the peripheral regions of the complex. Monomerization-induced loss of red chlorophylls is assigned to a cluster of chlorophylls adjacent to PsaX. Based on our findings, we propose a role of PsaX in the stabilization of red chlorophylls and that lipids of the surrounding membrane present a major source of thermal energy for uphill excitation energy transfer from red chlorophylls to P700.}, author = {Çoruh, Mehmet Orkun and Frank, Anna and Tanaka, Hideaki and Kawamoto, Akihiro and El-Mohsnawy, Eithar and Kato, Takayuki and Namba, Keiichi and Gerle, Christoph and Nowaczyk, Marc M. and Kurisu, Genji}, issn = {2399-3642}, journal = {Communications Biology}, keywords = {general agricultural and biological Sciences, general biochemistry, genetics and molecular biology, medicine (miscellaneous)}, number = {1}, publisher = {Springer }, title = {{Cryo-EM structure of a functional monomeric Photosystem I from Thermosynechococcus elongatus reveals red chlorophyll cluster}}, doi = {10.1038/s42003-021-01808-9}, volume = {4}, year = {2021}, } @article{10322, abstract = {To survive elevated temperatures, ectotherms adjust the fluidity of membranes by fine-tuning lipid desaturation levels in a process previously described to be cell autonomous. We have discovered that, in Caenorhabditis elegans, neuronal heat shock factor 1 (HSF-1), the conserved master regulator of the heat shock response (HSR), causes extensive fat remodeling in peripheral tissues. These changes include a decrease in fat desaturase and acid lipase expression in the intestine and a global shift in the saturation levels of plasma membrane’s phospholipids. The observed remodeling of plasma membrane is in line with ectothermic adaptive responses and gives worms a cumulative advantage to warm temperatures. We have determined that at least 6 TAX-2/TAX-4 cyclic guanosine monophosphate (cGMP) gated channel expressing sensory neurons, and transforming growth factor ß (TGF-β)/bone morphogenetic protein (BMP) are required for signaling across tissues to modulate fat desaturation. We also find neuronal hsf-1 is not only sufficient but also partially necessary to control the fat remodeling response and for survival at warm temperatures. This is the first study to show that a thermostat-based mechanism can cell nonautonomously coordinate membrane saturation and composition across tissues in a multicellular animal.}, author = {Chauve, Laetitia and Hodge, Francesca and Murdoch, Sharlene and Masoudzadeh, Fatemah and Mann, Harry Jack and Lopez-Clavijo, Andrea and Okkenhaug, Hanneke and West, Greg and Sousa, Bebiana C. and Segonds-Pichon, Anne and Li, Cheryl and Wingett, Steven and Kienberger, Hermine and Kleigrewe, Karin and De Bono, Mario and Wakelam, Michael and Casanueva, Olivia}, issn = {1545-7885}, journal = {PLoS Biology}, number = {11}, publisher = {Public Library of Science}, title = {{Neuronal HSF-1 coordinates the propagation of fat desaturation across tissues to enable adaptation to high temperatures in C. elegans}}, doi = {10.1371/journal.pbio.3001431}, volume = {19}, year = {2021}, } @article{10323, abstract = {Molecular chaperones are central to cellular protein homeostasis. Dynamic disorder is a key feature of the complexes of molecular chaperones and their client proteins, and it facilitates the client release towards a folded state or the handover to downstream components. The dynamic nature also implies that a given chaperone can interact with many different client proteins, based on physico-chemical sequence properties rather than on structural complementarity of their (folded) 3D structure. Yet, the balance between this promiscuity and some degree of client specificity is poorly understood. Here, we review recent atomic-level descriptions of chaperones with client proteins, including chaperones in complex with intrinsically disordered proteins, with membrane-protein precursors, or partially folded client proteins. We focus hereby on chaperone-client interactions that are independent of ATP. The picture emerging from these studies highlights the importance of dynamics in these complexes, whereby several interaction types, not only hydrophobic ones, contribute to the complex formation. We discuss these features of chaperone-client complexes and possible factors that may contribute to this balance of promiscuity and specificity.}, author = {Sučec, Iva and Bersch, Beate and Schanda, Paul}, issn = {2296-889X}, journal = {Frontiers in Molecular Biosciences}, publisher = {Frontiers}, title = {{How do chaperones bind (partly) unfolded client proteins?}}, doi = {10.3389/fmolb.2021.762005}, volume = {8}, year = {2021}, } @inproceedings{10324, abstract = {Off-chain protocols (channels) are a promising solution to the scalability and privacy challenges of blockchain payments. Current proposals, however, require synchrony assumptions to preserve the safety of a channel, leaking to an adversary the exact amount of time needed to control the network for a successful attack. In this paper, we introduce Brick, the first payment channel that remains secure under network asynchrony and concurrently provides correct incentives. The core idea is to incorporate the conflict resolution process within the channel by introducing a rational committee of external parties, called wardens. Hence, if a party wants to close a channel unilaterally, it can only get the committee’s approval for the last valid state. Additionally, Brick provides sub-second latency because it does not employ heavy-weight consensus. Instead, Brick uses consistent broadcast to announce updates and close the channel, a light-weight abstraction that is powerful enough to preserve safety and liveness to any rational parties. We formally define and prove for Brick the properties a payment channel construction should fulfill. We also design incentives for Brick such that honest and rational behavior aligns. Finally, we provide a reference implementation of the smart contracts in Solidity.}, author = {Avarikioti, Zeta and Kokoris Kogias, Eleftherios and Wattenhofer, Roger and Zindros, Dionysis}, booktitle = {25th International Conference on Financial Cryptography and Data Security}, isbn = {9-783-6626-4330-3}, issn = {1611-3349}, location = {Virtual}, pages = {209--230}, publisher = {Springer Nature}, title = {{Brick: Asynchronous incentive-compatible payment channels}}, doi = {10.1007/978-3-662-64331-0_11}, volume = {12675 }, year = {2021}, } @inproceedings{10325, abstract = {Since the inception of Bitcoin, a plethora of distributed ledgers differing in design and purpose has been created. While by design, blockchains provide no means to securely communicate with external systems, numerous attempts towards trustless cross-chain communication have been proposed over the years. Today, cross-chain communication (CCC) plays a fundamental role in cryptocurrency exchanges, scalability efforts via sharding, extension of existing systems through sidechains, and bootstrapping of new blockchains. Unfortunately, existing proposals are designed ad-hoc for specific use-cases, making it hard to gain confidence in their correctness and composability. We provide the first systematic exposition of cross-chain communication protocols. We formalize the underlying research problem and show that CCC is impossible without a trusted third party, contrary to common beliefs in the blockchain community. With this result in mind, we develop a framework to design new and evaluate existing CCC protocols, focusing on the inherent trust assumptions thereof, and derive a classification covering the field of cross-chain communication to date. We conclude by discussing open challenges for CCC research and the implications of interoperability on the security and privacy of blockchains.}, author = {Zamyatin, Alexei and Al-Bassam, Mustafa and Zindros, Dionysis and Kokoris Kogias, Eleftherios and Moreno-Sanchez, Pedro and Kiayias, Aggelos and Knottenbelt, William J.}, booktitle = {25th International Conference on Financial Cryptography and Data Security}, isbn = {9-783-6626-4330-3}, issn = {1611-3349}, location = {Virtual}, pages = {3--36}, publisher = {Springer Nature}, title = {{SoK: Communication across distributed ledgers}}, doi = {10.1007/978-3-662-64331-0_1}, volume = {12675 }, year = {2021}, } @article{10326, abstract = {Strigolactones (SLs) are carotenoid-derived plant hormones that control shoot branching and communications between host plants and symbiotic fungi or root parasitic plants. Extensive studies have identified the key components participating in SL biosynthesis and signalling, whereas the catabolism or deactivation of endogenous SLs in planta remains largely unknown. Here, we report that the Arabidopsis carboxylesterase 15 (AtCXE15) and its orthologues function as efficient hydrolases of SLs. We show that overexpression of AtCXE15 promotes shoot branching by dampening SL-inhibited axillary bud outgrowth. We further demonstrate that AtCXE15 could bind and efficiently hydrolyse SLs both in vitro and in planta. We also provide evidence that AtCXE15 is capable of catalysing hydrolysis of diverse SL analogues and that such CXE15-dependent catabolism of SLs is evolutionarily conserved in seed plants. These results disclose a catalytic mechanism underlying homoeostatic regulation of SLs in plants, which also provides a rational approach to spatial-temporally manipulate the endogenous SLs and thus architecture of crops and ornamental plants.}, author = {Xu, Enjun and Chai, Liang and Zhang, Shiqi and Yu, Ruixue and Zhang, Xixi and Xu, Chongyi and Hu, Yuxin}, issn = {2055-0278}, journal = {Nature Plants}, pages = {1495–1504 }, publisher = {Springer Nature}, title = {{Catabolism of strigolactones by a carboxylesterase}}, doi = {10.1038/s41477-021-01011-y}, volume = {7}, year = {2021}, } @article{10327, abstract = {Composite materials offer numerous advantages in a wide range of applications, including thermoelectrics. Here, semiconductor–metal composites are produced by just blending nanoparticles of a sulfide semiconductor obtained in aqueous solution and at room temperature with a metallic Cu powder. The obtained blend is annealed in a reducing atmosphere and afterward consolidated into dense polycrystalline pellets through spark plasma sintering (SPS). We observe that, during the annealing process, the presence of metallic copper activates a partial reduction of the PbS, resulting in the formation of PbS–Pb–CuxS composites. The presence of metallic lead during the SPS process habilitates the liquid-phase sintering of the composite. Besides, by comparing the transport properties of PbS, the PbS–Pb–CuxS composites, and PbS–CuxS composites obtained by blending PbS and CuxS nanoparticles, we demonstrate that the presence of metallic lead decisively contributes to a strong increase of the charge carrier concentration through spillover of charge carriers enabled by the low work function of lead. The increase in charge carrier concentration translates into much higher electrical conductivities and moderately lower Seebeck coefficients. These properties translate into power factors up to 2.1 mW m–1 K–2 at ambient temperature, well above those of PbS and PbS + CuxS. Additionally, the presence of multiple phases in the final composite results in a notable decrease in the lattice thermal conductivity. Overall, the introduction of metallic copper in the initial blend results in a significant improvement of the thermoelectric performance of PbS, reaching a dimensionless thermoelectric figure of merit ZT = 1.1 at 750 K, which represents about a 400% increase over bare PbS. Besides, an average ZTave = 0.72 in the temperature range 320–773 K is demonstrated.}, author = {Li, Mengyao and Liu, Yu and Zhang, Yu and Han, Xu and Xiao, Ke and Nabahat, Mehran and Arbiol, Jordi and Llorca, Jordi and Ibáñez, Maria and Cabot, Andreu}, issn = {1944-8252}, journal = {ACS Applied Materials and Interfaces}, keywords = {CuxS, PbS, energy conversion, nanocomposite, nanoparticle, solution synthesis, thermoelectric}, number = {43}, pages = {51373–51382}, publisher = {American Chemical Society }, title = {{PbS–Pb–CuxS composites for thermoelectric application}}, doi = {10.1021/acsami.1c15609}, volume = {13}, year = {2021}, } @article{10363, abstract = {Erythropoietin enhances oxygen delivery and reduces hypoxia-induced cell death, but its pro-thrombotic activity is problematic for use of erythropoietin in treating hypoxia. We constructed a fusion protein that stimulates red blood cell production and neuroprotection without triggering platelet production, a marker for thrombosis. The protein consists of an anti-glycophorin A nanobody and an erythropoietin mutant (L108A). The mutation reduces activation of erythropoietin receptor homodimers that induce erythropoiesis and thrombosis, but maintains the tissue-protective signaling. The binding of the nanobody element to glycophorin A rescues homodimeric erythropoietin receptor activation on red blood cell precursors. In a cell proliferation assay, the fusion protein is active at 10−14 M, allowing an estimate of the number of receptor–ligand complexes needed for signaling. This fusion protein stimulates erythroid cell proliferation in vitro and in mice, and shows neuroprotective activity in vitro. Our erythropoietin fusion protein presents a novel molecule for treating hypoxia.}, author = {Lee, Jungmin and Vernet, Andyna and Gruber, Nathalie and Kready, Kasia M. and Burrill, Devin R. and Way, Jeffrey C. and Silver, Pamela A.}, issn = {1741-0134}, journal = {Protein Engineering, Design and Selection}, publisher = {Oxford University Press}, title = {{Rational engineering of an erythropoietin fusion protein to treat hypoxia}}, doi = {10.1093/protein/gzab025}, volume = {34}, year = {2021}, } @article{10365, abstract = {The early development of many organisms involves the folding of cell monolayers, but this behaviour is difficult to reproduce in vitro; therefore, both mechanistic causes and effects of local curvature remain unclear. Here we study epithelial cell monolayers on corrugated hydrogels engineered into wavy patterns, examining how concave and convex curvatures affect cellular and nuclear shape. We find that substrate curvature affects monolayer thickness, which is larger in valleys than crests. We show that this feature generically arises in a vertex model, leading to the hypothesis that cells may sense curvature by modifying the thickness of the tissue. We find that local curvature also affects nuclear morphology and positioning, which we explain by extending the vertex model to take into account membrane–nucleus interactions, encoding thickness modulation in changes to nuclear deformation and position. We propose that curvature governs the spatial distribution of yes-associated proteins via nuclear shape and density changes. We show that curvature also induces significant variations in lamins, chromatin condensation and cell proliferation rate in folded epithelial tissues. Together, this work identifies active cell mechanics and nuclear mechanoadaptation as the key players of the mechanistic regulation of epithelia to substrate curvature.}, author = {Luciano, Marine and Xue, Shi-lei and De Vos, Winnok H. and Redondo-Morata, Lorena and Surin, Mathieu and Lafont, Frank and Hannezo, Edouard B and Gabriele, Sylvain}, issn = {1745-2481}, journal = {Nature Physics}, number = {12}, pages = {1382–1390}, publisher = {Springer Nature}, title = {{Cell monolayers sense curvature by exploiting active mechanics and nuclear mechanoadaptation}}, doi = {10.1038/s41567-021-01374-1}, volume = {17}, year = {2021}, } @article{10366, author = {Heisenberg, Carl-Philipp J and Lennon, Ana Maria and Mayor, Roberto and Salbreux, Guillaume}, issn = {2667-2901}, journal = {Cells and Development}, number = {12}, publisher = {Elsevier}, title = {{Special rebranding issue: “Quantitative cell and developmental biology”}}, doi = {10.1016/j.cdev.2021.203758}, volume = {168}, year = {2021}, } @inproceedings{10367, abstract = {How information is created, shared and consumed has changed rapidly in recent decades, in part thanks to new social platforms and technologies on the web. With ever-larger amounts of unstructured and limited labels, organizing and reconciling information from different sources and modalities is a central challenge in machine learning. This cutting-edge tutorial aims to introduce the multimodal entailment task, which can be useful for detecting semantic alignments when a single modality alone does not suffice for a whole content understanding. Starting with a brief overview of natural language processing, computer vision, structured data and neural graph learning, we lay the foundations for the multimodal sections to follow. We then discuss recent multimodal learning literature covering visual, audio and language streams, and explore case studies focusing on tasks which require fine-grained understanding of visual and linguistic semantics question answering, veracity and hatred classification. Finally, we introduce a new dataset for recognizing multimodal entailment, exploring it in a hands-on collaborative section. Overall, this tutorial gives an overview of multimodal learning, introduces a multimodal entailment dataset, and encourages future research in the topic.}, author = {Ilharco, Cesar and Shirazi, Afsaneh and Gopalan, Arjun and Nagrani, Arsha and Bratanič, Blaž and Bregler, Chris and Liu, Christina and Ferreira, Felipe and Barcik, Gabriek and Ilharco, Gabriel and Osang, Georg F and Bulian, Jannis and Frank, Jared and Smaira, Lucas and Cao, Qin and Marino, Ricardo and Patel, Roma and Leung, Thomas and Imbrasaite, Vaiva}, booktitle = {59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Tutorial Abstracts}, isbn = {9-781-9540-8557-2}, location = {Bangkok, Thailand}, pages = {29--30}, publisher = {Association for Computational Linguistics}, title = {{Recognizing multimodal entailment}}, doi = {10.18653/v1/2021.acl-tutorials.6}, year = {2021}, } @article{10402, abstract = {Branching morphogenesis governs the formation of many organs such as lung, kidney, and the neurovascular system. Many studies have explored system-specific molecular and cellular regulatory mechanisms, as well as self-organizing rules underlying branching morphogenesis. However, in addition to local cues, branched tissue growth can also be influenced by global guidance. Here, we develop a theoretical framework for a stochastic self-organized branching process in the presence of external cues. Combining analytical theory with numerical simulations, we predict differential signatures of global vs. local regulatory mechanisms on the branching pattern, such as angle distributions, domain size, and space-filling efficiency. We find that branch alignment follows a generic scaling law determined by the strength of global guidance, while local interactions influence the tissue density but not its overall territory. Finally, using zebrafish innervation as a model system, we test these key features of the model experimentally. Our work thus provides quantitative predictions to disentangle the role of different types of cues in shaping branched structures across scales.}, author = {Ucar, Mehmet C and Kamenev, Dmitrii and Sunadome, Kazunori and Fachet, Dominik C and Lallemend, Francois and Adameyko, Igor and Hadjab, Saida and Hannezo, Edouard B}, issn = {2041-1723}, journal = {Nature Communications}, publisher = {Springer Nature}, title = {{Theory of branching morphogenesis by local interactions and global guidance}}, doi = {10.1038/s41467-021-27135-5}, volume = {12}, year = {2021}, } @article{10403, abstract = {Synaptic transmission, connectivity, and dendritic morphology mature in parallel during brain development and are often disrupted in neurodevelopmental disorders. Yet how these changes influence the neuronal computations necessary for normal brain function are not well understood. To identify cellular mechanisms underlying the maturation of synaptic integration in interneurons, we combined patch-clamp recordings of excitatory inputs in mouse cerebellar stellate cells (SCs), three-dimensional reconstruction of SC morphology with excitatory synapse location, and biophysical modeling. We found that postnatal maturation of postsynaptic strength was homogeneously reduced along the somatodendritic axis, but dendritic integration was always sublinear. However, dendritic branching increased without changes in synapse density, leading to a substantial gain in distal inputs. Thus, changes in synapse distribution, rather than dendrite cable properties, are the dominant mechanism underlying the maturation of neuronal computation. These mechanisms favor the emergence of a spatially compartmentalized two-stage integration model promoting location-dependent integration within dendritic subunits.}, author = {Biane, Celia and Rückerl, Florian and Abrahamsson, Therese and Saint-Cloment, Cécile and Mariani, Jean and Shigemoto, Ryuichi and Digregorio, David A. and Sherrard, Rachel M. and Cathala, Laurence}, issn = {2050-084X}, journal = {eLife}, publisher = {eLife Sciences Publications}, title = {{Developmental emergence of two-stage nonlinear synaptic integration in cerebellar interneurons}}, doi = {10.7554/eLife.65954}, volume = {10}, year = {2021}, } @article{10404, abstract = {While convolutional neural networks (CNNs) have found wide adoption as state-of-the-art models for image-related tasks, their predictions are often highly sensitive to small input perturbations, which the human vision is robust against. This paper presents Perturber, a web-based application that allows users to instantaneously explore how CNN activations and predictions evolve when a 3D input scene is interactively perturbed. Perturber offers a large variety of scene modifications, such as camera controls, lighting and shading effects, background modifications, object morphing, as well as adversarial attacks, to facilitate the discovery of potential vulnerabilities. Fine-tuned model versions can be directly compared for qualitative evaluation of their robustness. Case studies with machine learning experts have shown that Perturber helps users to quickly generate hypotheses about model vulnerabilities and to qualitatively compare model behavior. Using quantitative analyses, we could replicate users’ insights with other CNN architectures and input images, yielding new insights about the vulnerability of adversarially trained models.}, author = {Sietzen, Stefan and Lechner, Mathias and Borowski, Judy and Hasani, Ramin and Waldner, Manuela}, issn = {1467-8659}, journal = {Computer Graphics Forum}, number = {7}, pages = {253--264}, publisher = {Wiley}, title = {{Interactive analysis of CNN robustness}}, doi = {10.1111/cgf.14418}, volume = {40}, year = {2021}, }