TY - JOUR AB - In the computation of the material properties of random alloys, the method of 'special quasirandom structures' attempts to approximate the properties of the alloy on a finite volume with higher accuracy by replicating certain statistics of the random atomic lattice in the finite volume as accurately as possible. In the present work, we provide a rigorous justification for a variant of this method in the framework of the Thomas–Fermi–von Weizsäcker (TFW) model. Our approach is based on a recent analysis of a related variance reduction method in stochastic homogenization of linear elliptic PDEs and the locality properties of the TFW model. Concerning the latter, we extend an exponential locality result by Nazar and Ortner to include point charges, a result that may be of independent interest. AU - Fischer, Julian L AU - Kniely, Michael ID - 8697 IS - 11 JF - Nonlinearity SN - 09517715 TI - Variance reduction for effective energies of random lattices in the Thomas-Fermi-von Weizsäcker model VL - 33 ER - TY - JOUR AB - Animal development entails the organization of specific cell types in space and time, and spatial patterns must form in a robust manner. In the zebrafish spinal cord, neural progenitors form stereotypic patterns despite noisy morphogen signaling and large-scale cellular rearrangements during morphogenesis and growth. By directly measuring adhesion forces and preferences for three types of endogenous neural progenitors, we provide evidence for the differential adhesion model in which differences in intercellular adhesion mediate cell sorting. Cell type–specific combinatorial expression of different classes of cadherins (N-cadherin, cadherin 11, and protocadherin 19) results in homotypic preference ex vivo and patterning robustness in vivo. Furthermore, the differential adhesion code is regulated by the sonic hedgehog morphogen gradient. We propose that robust patterning during tissue morphogenesis results from interplay between adhesion-based self-organization and morphogen-directed patterning. AU - Tsai, Tony Y.-C. AU - Sikora, Mateusz K AU - Xia, Peng AU - Colak-Champollion, Tugba AU - Knaut, Holger AU - Heisenberg, Carl-Philipp J AU - Megason, Sean G. ID - 8680 IS - 6512 JF - Science KW - Multidisciplinary SN - 0036-8075 TI - An adhesion code ensures robust pattern formation during tissue morphogenesis VL - 370 ER - TY - JOUR AB - Dynamic changes in the three-dimensional (3D) organization of chromatin are associated with central biological processes, such as transcription, replication and development. Therefore, the comprehensive identification and quantification of these changes is fundamental to understanding of evolutionary and regulatory mechanisms. Here, we present Comparison of Hi-C Experiments using Structural Similarity (CHESS), an algorithm for the comparison of chromatin contact maps and automatic differential feature extraction. We demonstrate the robustness of CHESS to experimental variability and showcase its biological applications on (1) interspecies comparisons of syntenic regions in human and mouse models; (2) intraspecies identification of conformational changes in Zelda-depleted Drosophila embryos; (3) patient-specific aberrant chromatin conformation in a diffuse large B-cell lymphoma sample; and (4) the systematic identification of chromatin contact differences in high-resolution Capture-C data. In summary, CHESS is a computationally efficient method for the comparison and classification of changes in chromatin contact data. AU - Galan, Silvia AU - Machnik, Nick N AU - Kruse, Kai AU - Díaz, Noelia AU - Marti-Renom, Marc A AU - Vaquerizas, Juan M ID - 8707 JF - Nature Genetics SN - 10614036 TI - CHESS enables quantitative comparison of chromatin contact data and automatic feature extraction VL - 52 ER - TY - JOUR AB - A central goal of artificial intelligence in high-stakes decision-making applications is to design a single algorithm that simultaneously expresses generalizability by learning coherent representations of their world and interpretable explanations of its dynamics. Here, we combine brain-inspired neural computation principles and scalable deep learning architectures to design compact neural controllers for task-specific compartments of a full-stack autonomous vehicle control system. We discover that a single algorithm with 19 control neurons, connecting 32 encapsulated input features to outputs by 253 synapses, learns to map high-dimensional inputs into steering commands. This system shows superior generalizability, interpretability and robustness compared with orders-of-magnitude larger black-box learning systems. The obtained neural agents enable high-fidelity autonomy for task-specific parts of a complex autonomous system. AU - Lechner, Mathias AU - Hasani, Ramin AU - Amini, Alexander AU - Henzinger, Thomas A AU - Rus, Daniela AU - Grosu, Radu ID - 8679 JF - Nature Machine Intelligence TI - Neural circuit policies enabling auditable autonomy VL - 2 ER - TY - JOUR AB - The α–z Rényi relative entropies are a two-parameter family of Rényi relative entropies that are quantum generalizations of the classical α-Rényi relative entropies. In the work [Adv. Math. 365, 107053 (2020)], we decided the full range of (α, z) for which the data processing inequality (DPI) is valid. In this paper, we give algebraic conditions for the equality in DPI. For the full range of parameters (α, z), we give necessary conditions and sufficient conditions. For most parameters, we give equivalent conditions. This generalizes and strengthens the results of Leditzky et al. [Lett. Math. Phys. 107, 61–80 (2017)]. AU - Zhang, Haonan ID - 8670 IS - 10 JF - Journal of Mathematical Physics SN - 00222488 TI - Equality conditions of data processing inequality for α-z Rényi relative entropies VL - 61 ER - TY - JOUR AB - The brain represents and reasons probabilistically about complex stimuli and motor actions using a noisy, spike-based neural code. A key building block for such neural computations, as well as the basis for supervised and unsupervised learning, is the ability to estimate the surprise or likelihood of incoming high-dimensional neural activity patterns. Despite progress in statistical modeling of neural responses and deep learning, current approaches either do not scale to large neural populations or cannot be implemented using biologically realistic mechanisms. Inspired by the sparse and random connectivity of real neuronal circuits, we present a model for neural codes that accurately estimates the likelihood of individual spiking patterns and has a straightforward, scalable, efficient, learnable, and realistic neural implementation. This model’s performance on simultaneously recorded spiking activity of >100 neurons in the monkey visual and prefrontal cortices is comparable with or better than that of state-of-the-art models. Importantly, the model can be learned using a small number of samples and using a local learning rule that utilizes noise intrinsic to neural circuits. Slower, structural changes in random connectivity, consistent with rewiring and pruning processes, further improve the efficiency and sparseness of the resulting neural representations. Our results merge insights from neuroanatomy, machine learning, and theoretical neuroscience to suggest random sparse connectivity as a key design principle for neuronal computation. AU - Maoz, Ori AU - Tkačik, Gašper AU - Esteki, Mohamad Saleh AU - Kiani, Roozbeh AU - Schneidman, Elad ID - 8698 IS - 40 JF - Proceedings of the National Academy of Sciences of the United States of America SN - 00278424 TI - Learning probabilistic neural representations with randomly connected circuits VL - 117 ER - TY - CONF AB - Traditional robotic control suits require profound task-specific knowledge for designing, building and testing control software. The rise of Deep Learning has enabled end-to-end solutions to be learned entirely from data, requiring minimal knowledge about the application area. We design a learning scheme to train end-to-end linear dynamical systems (LDS)s by gradient descent in imitation learning robotic domains. We introduce a new regularization loss component together with a learning algorithm that improves the stability of the learned autonomous system, by forcing the eigenvalues of the internal state updates of an LDS to be negative reals. We evaluate our approach on a series of real-life and simulated robotic experiments, in comparison to linear and nonlinear Recurrent Neural Network (RNN) architectures. Our results show that our stabilizing method significantly improves test performance of LDS, enabling such linear models to match the performance of contemporary nonlinear RNN architectures. A video of the obstacle avoidance performance of our method on a mobile robot, in unseen environments, compared to other methods can be viewed at https://youtu.be/mhEsCoNao5E. AU - Lechner, Mathias AU - Hasani, Ramin AU - Rus, Daniela AU - Grosu, Radu ID - 8704 SN - 10504729 T2 - Proceedings - IEEE International Conference on Robotics and Automation TI - Gershgorin loss stabilizes the recurrent neural network compartment of an end-to-end robot learning scheme ER - TY - JOUR AB - Translation termination is a finishing step of protein biosynthesis. The significant role in this process belongs not only to protein factors of translation termination but also to the nearest nucleotide environment of stop codons. There are numerous descriptions of stop codons readthrough, which is due to specific nucleotide sequences behind them. However, represented data are segmental and don’t explain the mechanism of the nucleotide context influence on translation termination. It is well known that stop codon UAA usage is preferential for A/T-rich genes, and UAG, UGA—for G/C-rich genes, which is related to an expression level of these genes. We investigated the connection between a frequency of nucleotides occurrence in 3' area of stop codons in the human genome and their influence on translation termination efficiency. We found that 3' context motif, which is cognate to the sequence of a stop codon, stimulates translation termination. At the same time, the nucleotide composition of 3' sequence that differs from stop codon, decreases translation termination efficiency. AU - Sokolova, E. E. AU - Vlasov, Petr AU - Egorova, T. V. AU - Shuvalov, A. V. AU - Alkalaeva, E. Z. ID - 8700 IS - 5 JF - Molecular Biology SN - 00268933 TI - The influence of A/G composition of 3' stop codon contexts on translation termination efficiency in eukaryotes VL - 54 ER - TY - JOUR AB - Translation termination is a finishing step of protein biosynthesis. The significant role in this process belongs not only to protein factors of translation termination but also to the nearest nucleotide environment of stop codons. There are numerous descriptions of stop codons readthrough, which is due to specific nucleotide sequences behind them. However, represented data are segmental and don’t explain the mechanism of the nucleotide context influence on translation termination. It is well known that stop codon UAA usage is preferential for A/T-rich genes, and UAG, UGA—for G/C-rich genes, which is related to an expression level of these genes. We investigated the connection between a frequency of nucleotides occurrence in 3' area of stop codons in the human genome and their influence on translation termination efficiency. We found that 3' context motif, which is cognate to the sequence of a stop codon, stimulates translation termination. At the same time, the nucleotide composition of 3' sequence that differs from stop codon, decreases translation termination efficiency. AU - Sokolova, E. E. AU - Vlasov, Petr AU - Egorova, T. V. AU - Shuvalov, A. V. AU - Alkalaeva, E. Z. ID - 8701 IS - 5 JF - Molekuliarnaia biologiia SN - 00268984 TI - The influence of A/G composition of 3' stop codon contexts on translation termination efficiency in eukaryotes VL - 54 ER - TY - GEN AB - A binary neutron star merger has been observed in a multi-messenger detection of gravitational wave (GW) and electromagnetic (EM) radiation. Binary neutron stars that merge within a Hubble time, as well as many other compact binaries, are expected to form via common envelope evolution. Yet five decades of research on common envelope evolution have not yet resulted in a satisfactory understanding of the multi-spatial multi-timescale evolution for the systems that lead to compact binaries. In this paper, we report on the first successful simulations of common envelope ejection leading to binary neutron star formation in 3D hydrodynamics. We simulate the dynamical inspiral phase of the interaction between a 12M⊙ red supergiant and a 1.4M⊙ neutron star for different initial separations and initial conditions. For all of our simulations, we find complete envelope ejection and final orbital separations of af≈1.3-5.1R⊙ depending on the simulation and criterion, leading to binary neutron stars that can merge within a Hubble time. We find αCE-equivalent efficiencies of ≈0.1-2.7 depending on the simulation and criterion, but this may be specific for these extended progenitors. We fully resolve the core of the star to ≲0.005R⊙ and our 3D hydrodynamics simulations are informed by an adjusted 1D analytic energy formalism and a 2D kinematics study in order to overcome the prohibitive computational cost of simulating these systems. The framework we develop in this paper can be used to simulate a wide variety of interactions between stars, from stellar mergers to common envelope episodes leading to GW sources. AU - Jamie A. P. Law-Smith, Jamie A. P. Law-Smith AU - Everson, Rosa Wallace AU - Enrico Ramirez-Ruiz, Enrico Ramirez-Ruiz AU - Mink, Selma E. de AU - Son, Lieke A. C. van AU - Götberg, Ylva Louise Linsdotter AU - Zellmann, Stefan AU - Alejandro Vigna-Gómez, Alejandro Vigna-Gómez AU - Renzo, Mathieu AU - Wu, Samantha AU - Schrøder, Sophie L. AU - Foley, Ryan J. AU - Tenley Hutchinson-Smith, Tenley Hutchinson-Smith ID - 14096 T2 - arXiv TI - Successful common envelope ejection and binary neutron star formation in 3D hydrodynamics ER -