Cell polarization is a critical process that separates molecules into two distinct regions in prokaryotic and eukaryotic cells, guiding biological processes such as cell division and cell differentiation. Although several underlying antagonistic reaction-diffusion networks capable of setting up cell polarization have been identified experimentally and theoretically, our understanding of how to manipulate pattern stability and asymmetry remains incomplete, especially when only a subset of network components are known. Here we present numerical results to show that the polarized pattern of an antagonistic 2-node network collapses into a homogeneous state when subjected to single-sided self-regulation, single-sided additional regulation, or unequal system parameters. However, polarity can be restored through a combination of two modifications that have opposing effects. Additionally, spatially inhomogeneous parameters favoring respective domains stabilize their interface at designated locations. To connect our findings to cell polarity studies of the nematode Caenorhabditis elegans zygote, we reconstituted a 5-node network where a 4-node circuit with full mutual inhibitions between anterior and posterior is modified by a mutual activation in the anterior and an additional mutual inhibition between the anterior and the posterior. Once again, a generic set of kinetic parameters moves the interface towards either the anterior or posterior end, yet a polarized pattern can be stabilized through spatial tuning of one or more parameters coupled to intracellular or extracellular cues. A user-friendly software, PolarSim, is introduced to facilitate the exploration of networks with alternative node numbers, parameter values, and regulatory pathways.
{"title":"Balancing reaction-diffusion network for cell polarization pattern with stability and asymmetry","authors":"Yixuan Chen, Guoye Guan, Lei-Han Tang, Chao Tang","doi":"arxiv-2401.07227","DOIUrl":"https://doi.org/arxiv-2401.07227","url":null,"abstract":"Cell polarization is a critical process that separates molecules into two\u0000distinct regions in prokaryotic and eukaryotic cells, guiding biological\u0000processes such as cell division and cell differentiation. Although several\u0000underlying antagonistic reaction-diffusion networks capable of setting up cell\u0000polarization have been identified experimentally and theoretically, our\u0000understanding of how to manipulate pattern stability and asymmetry remains\u0000incomplete, especially when only a subset of network components are known. Here\u0000we present numerical results to show that the polarized pattern of an\u0000antagonistic 2-node network collapses into a homogeneous state when subjected\u0000to single-sided self-regulation, single-sided additional regulation, or unequal\u0000system parameters. However, polarity can be restored through a combination of\u0000two modifications that have opposing effects. Additionally, spatially\u0000inhomogeneous parameters favoring respective domains stabilize their interface\u0000at designated locations. To connect our findings to cell polarity studies of\u0000the nematode Caenorhabditis elegans zygote, we reconstituted a 5-node network\u0000where a 4-node circuit with full mutual inhibitions between anterior and\u0000posterior is modified by a mutual activation in the anterior and an additional\u0000mutual inhibition between the anterior and the posterior. Once again, a generic\u0000set of kinetic parameters moves the interface towards either the anterior or\u0000posterior end, yet a polarized pattern can be stabilized through spatial tuning\u0000of one or more parameters coupled to intracellular or extracellular cues. A\u0000user-friendly software, PolarSim, is introduced to facilitate the exploration\u0000of networks with alternative node numbers, parameter values, and regulatory\u0000pathways.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139481682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dimitri Loutchko, Yuki Sughiyama, Tetsuya J. Kobayashi
Chemical reaction networks (CRN) comprise an important class of models to understand biological functions such as cellular information processing, the robustness and control of metabolic pathways, circadian rhythms, and many more. However, any CRN describing a certain function does not act in isolation but is a part of a much larger network and as such is constantly subject to external changes. In [Shinar, Alon, and Feinberg. "Sensitivity and robustness in chemical reaction networks." SIAM J App Math (2009): 977-998.], the responses of CRN to changes in the linear conserved quantities, called sensitivities, were studied in and the question of how to construct absolute, i.e., basis-independent, sensitivities was raised. In this article, by applying information geometric methods, such a construction is provided. The idea is to track how concentration changes in a particular chemical propagate to changes of all the other chemicals within a steady state. This is encoded in the matrix of absolute sensitivites. A linear algebraic characterization of the matrix of absolute sensitivities for quasi-thermostatic CRN is derived via a Cramer-Rao bound for CRN, which is based on the the analogy between quasi-thermostatic steady states and the exponential family of probability distributions.
{"title":"Cramer-Rao bound and absolute sensitivity in chemical reaction networks","authors":"Dimitri Loutchko, Yuki Sughiyama, Tetsuya J. Kobayashi","doi":"arxiv-2401.06987","DOIUrl":"https://doi.org/arxiv-2401.06987","url":null,"abstract":"Chemical reaction networks (CRN) comprise an important class of models to\u0000understand biological functions such as cellular information processing, the\u0000robustness and control of metabolic pathways, circadian rhythms, and many more.\u0000However, any CRN describing a certain function does not act in isolation but is\u0000a part of a much larger network and as such is constantly subject to external\u0000changes. In [Shinar, Alon, and Feinberg. \"Sensitivity and robustness in\u0000chemical reaction networks.\" SIAM J App Math (2009): 977-998.], the responses\u0000of CRN to changes in the linear conserved quantities, called sensitivities,\u0000were studied in and the question of how to construct absolute, i.e.,\u0000basis-independent, sensitivities was raised. In this article, by applying\u0000information geometric methods, such a construction is provided. The idea is to\u0000track how concentration changes in a particular chemical propagate to changes\u0000of all the other chemicals within a steady state. This is encoded in the matrix\u0000of absolute sensitivites. A linear algebraic characterization of the matrix of\u0000absolute sensitivities for quasi-thermostatic CRN is derived via a Cramer-Rao\u0000bound for CRN, which is based on the the analogy between quasi-thermostatic\u0000steady states and the exponential family of probability distributions.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139484168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recent years have seen a rapid growth of machine learning in cheminformatics problems. In order to tackle the problem of insufficient training data in reality, more and more researchers pay attention to data augmentation technology. However, few researchers pay attention to the problem of construction rules and domain information of data, which will directly impact the quality of augmented data and the augmentation performance. While in graph-based molecular research, the molecular connectivity index, as a critical topological index, can directly or indirectly reflect the topology-based physicochemical properties and biological activities. In this paper, we propose a novel data augmentation technique that modifies the topology of the molecular graph to generate augmented data with the same molecular connectivity index as the original data. The molecular connectivity index combined with data augmentation technology helps to retain more topology-based molecular properties information and generate more reliable data. Furthermore, we adopt five benchmark datasets to test our proposed models, and the results indicate that the augmented data generated based on important molecular topology features can effectively improve the prediction accuracy of molecular properties, which also provides a new perspective on data augmentation in cheminformatics studies.
{"title":"Multi-Modal Representation Learning for Molecular Property Prediction: Sequence, Graph, Geometry","authors":"Zeyu Wang, Tianyi Jiang, Jinhuan Wang, Qi Xuan","doi":"arxiv-2401.03369","DOIUrl":"https://doi.org/arxiv-2401.03369","url":null,"abstract":"Recent years have seen a rapid growth of machine learning in cheminformatics\u0000problems. In order to tackle the problem of insufficient training data in\u0000reality, more and more researchers pay attention to data augmentation\u0000technology. However, few researchers pay attention to the problem of\u0000construction rules and domain information of data, which will directly impact\u0000the quality of augmented data and the augmentation performance. While in\u0000graph-based molecular research, the molecular connectivity index, as a critical\u0000topological index, can directly or indirectly reflect the topology-based\u0000physicochemical properties and biological activities. In this paper, we propose\u0000a novel data augmentation technique that modifies the topology of the molecular\u0000graph to generate augmented data with the same molecular connectivity index as\u0000the original data. The molecular connectivity index combined with data\u0000augmentation technology helps to retain more topology-based molecular\u0000properties information and generate more reliable data. Furthermore, we adopt\u0000five benchmark datasets to test our proposed models, and the results indicate\u0000that the augmented data generated based on important molecular topology\u0000features can effectively improve the prediction accuracy of molecular\u0000properties, which also provides a new perspective on data augmentation in\u0000cheminformatics studies.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139413777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Embedding sequential computations in biochemical environments is challenging because the computations are carried out by chemical reactions, which are inherently disordered. In this paper we apply modular design to specific calculations through chemical reactions and provide a design scheme of biochemical oscillator models in order to generate periodical species for the order regulation of these reaction modules. We take the case of arbitrary multi-module regulation into consideration, analyze the main errors in the regulation process under textit{mass-action kinetics} and demonstrate our design scheme under existing synthetic biochemical oscillator models.
{"title":"Controlling the occurrence sequence of reaction modules through biochemical relaxation oscillators","authors":"Xiaopeng Shi, Chuanhou Gao, Denis Dochain","doi":"arxiv-2401.02061","DOIUrl":"https://doi.org/arxiv-2401.02061","url":null,"abstract":"Embedding sequential computations in biochemical environments is challenging\u0000because the computations are carried out by chemical reactions, which are\u0000inherently disordered. In this paper we apply modular design to specific\u0000calculations through chemical reactions and provide a design scheme of\u0000biochemical oscillator models in order to generate periodical species for the\u0000order regulation of these reaction modules. We take the case of arbitrary\u0000multi-module regulation into consideration, analyze the main errors in the\u0000regulation process under textit{mass-action kinetics} and demonstrate our\u0000design scheme under existing synthetic biochemical oscillator models.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139102209","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Luis David García Puente, Elizabeth Gross, Heather A Harrington, Matthew Johnston, Nicolette Meshkat, Mercedes Pérez Millán, Anne Shiu
Motivated by the question of how biological systems maintain homeostasis in changing environments, Shinar and Feinberg introduced in 2010 the concept of absolute concentration robustness (ACR). A biochemical system exhibits ACR in some species if the steady-state value of that species does not depend on initial conditions. Thus, a system with ACR can maintain a constant level of one species even as the environment changes. Despite a great deal of interest in ACR in recent years, the following basic question remains open: How can we determine quickly whether a given biochemical system has ACR? Although various approaches to this problem have been proposed, we show that they are incomplete. Accordingly, we present new methods for deciding ACR, which harness computational algebra. We illustrate our results on several biochemical signaling networks.
{"title":"Absolute concentration robustness: Algebra and geometry","authors":"Luis David García Puente, Elizabeth Gross, Heather A Harrington, Matthew Johnston, Nicolette Meshkat, Mercedes Pérez Millán, Anne Shiu","doi":"arxiv-2401.00078","DOIUrl":"https://doi.org/arxiv-2401.00078","url":null,"abstract":"Motivated by the question of how biological systems maintain homeostasis in\u0000changing environments, Shinar and Feinberg introduced in 2010 the concept of\u0000absolute concentration robustness (ACR). A biochemical system exhibits ACR in\u0000some species if the steady-state value of that species does not depend on\u0000initial conditions. Thus, a system with ACR can maintain a constant level of\u0000one species even as the environment changes. Despite a great deal of interest\u0000in ACR in recent years, the following basic question remains open: How can we\u0000determine quickly whether a given biochemical system has ACR? Although various\u0000approaches to this problem have been proposed, we show that they are\u0000incomplete. Accordingly, we present new methods for deciding ACR, which harness\u0000computational algebra. We illustrate our results on several biochemical\u0000signaling networks.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139077930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The Escherichia coli chemoreceptors form an extensive array that achieves cooperative and adaptive sensing of extracellular signals. The receptors control the activity of histidine kinase CheA, which drives a non-equilibrium phosphorylation-dephosphorylation reaction cycle for response regulator CheY. Recent single-cell FRET measurements revealed that kinase activity of the array spontaneously switches between active and inactive states, with asymmetric switching times that signify time-reversal symmetry breaking in the underlying dynamics. Here, we show that the asymmetric switching dynamics can be explained by a non-equilibrium lattice model, which considers both the dissipative reaction cycles of individual core units and the coupling between neighboring units. The model reveals that large dissipation and near-critical coupling are required to explain the observed switching dynamics. Microscopically, the switching time asymmetry originates from irreversible transition paths. The model shows that strong dissipation enables sensitive and rapid signaling response by relieving the speed-sensitivity trade-off, which can be tested by future single-cell experiments. Overall, our model provides a general framework for studying biological complexes composed of coupled subunits that are individually driven by dissipative cycles and the rich non-equilibrium physics within.
{"title":"Time-reversal symmetry breaking in the chemosensory array: asymmetric switching and dissipation-enhanced sensing","authors":"David Hathcock, Qiwei Yu, Yuhai Tu","doi":"arxiv-2312.17424","DOIUrl":"https://doi.org/arxiv-2312.17424","url":null,"abstract":"The Escherichia coli chemoreceptors form an extensive array that achieves\u0000cooperative and adaptive sensing of extracellular signals. The receptors\u0000control the activity of histidine kinase CheA, which drives a non-equilibrium\u0000phosphorylation-dephosphorylation reaction cycle for response regulator CheY.\u0000Recent single-cell FRET measurements revealed that kinase activity of the array\u0000spontaneously switches between active and inactive states, with asymmetric\u0000switching times that signify time-reversal symmetry breaking in the underlying\u0000dynamics. Here, we show that the asymmetric switching dynamics can be explained\u0000by a non-equilibrium lattice model, which considers both the dissipative\u0000reaction cycles of individual core units and the coupling between neighboring\u0000units. The model reveals that large dissipation and near-critical coupling are\u0000required to explain the observed switching dynamics. Microscopically, the\u0000switching time asymmetry originates from irreversible transition paths. The\u0000model shows that strong dissipation enables sensitive and rapid signaling\u0000response by relieving the speed-sensitivity trade-off, which can be tested by\u0000future single-cell experiments. Overall, our model provides a general framework\u0000for studying biological complexes composed of coupled subunits that are\u0000individually driven by dissipative cycles and the rich non-equilibrium physics\u0000within.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139071643","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Miloš Nikolić, Victoria Antonetti, Feng Liu, Gentian Muhaxheri, Mariela D. Petkova, Martin Scheeler, Eric M. Smith, William Bialek, Thomas Gregor
The body plan of the fruit fly is determined by the expression of just a handful of genes. We show that the spatial patterns of expression for several of these genes scale precisely with the size of the embryo. Concretely, discrete positional markers such as the peaks in striped patterns have absolute positions along the anterior-posterior axis that are proportional to embryo length, with better than 1% accuracy. Further, the information (in bits) that graded patterns of expression provide about position can be decomposed into information about fractional or scaled position and information about absolute position or embryo length; all of the available information is about scaled position, again with ~1% accuracy. These observations suggest that the underlying genetic network exhibits scale invariance in a deeper mathematical sense. Taking this mathematical statement seriously requires that the network dynamics have a zero mode, which connects to many other observations on this system.
{"title":"Scale invariance in early embryonic development","authors":"Miloš Nikolić, Victoria Antonetti, Feng Liu, Gentian Muhaxheri, Mariela D. Petkova, Martin Scheeler, Eric M. Smith, William Bialek, Thomas Gregor","doi":"arxiv-2312.17684","DOIUrl":"https://doi.org/arxiv-2312.17684","url":null,"abstract":"The body plan of the fruit fly is determined by the expression of just a\u0000handful of genes. We show that the spatial patterns of expression for several\u0000of these genes scale precisely with the size of the embryo. Concretely,\u0000discrete positional markers such as the peaks in striped patterns have absolute\u0000positions along the anterior-posterior axis that are proportional to embryo\u0000length, with better than 1% accuracy. Further, the information (in bits) that\u0000graded patterns of expression provide about position can be decomposed into\u0000information about fractional or scaled position and information about absolute\u0000position or embryo length; all of the available information is about scaled\u0000position, again with ~1% accuracy. These observations suggest that the\u0000underlying genetic network exhibits scale invariance in a deeper mathematical\u0000sense. Taking this mathematical statement seriously requires that the network\u0000dynamics have a zero mode, which connects to many other observations on this\u0000system.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139071714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The problem of evolutionary complexification of life is considered one of the fundamental aspects in contemporary evolutionary theory. Parasitism is ubiquitous, inevitable, and arises as soon as the first replicators appear, even during the prebiotic stages of evolution. Both in theoretical approaches (computer modeling and analysis) and in real experiments (replication of biological macromolecules), parasitic processes emerge almost immediately. An effective way to avoid the elimination of the host-parasite system is through compartmentalization. In both theory and experiments, the pressure of parasitism leads to the complexification of the host-parasite system into a network of cooperative replicators and their parasites. Parasites have the ability to create niches for new replicators. The co-evolutionary arms race between defense systems and counter-defense mechanisms among parasites and hosts can progress for a considerable duration, involving multiple stages, if not indefinitely.
{"title":"Co-evolution of replicators and their parasites","authors":"Alexander Spirov","doi":"arxiv-2312.17540","DOIUrl":"https://doi.org/arxiv-2312.17540","url":null,"abstract":"The problem of evolutionary complexification of life is considered one of the\u0000fundamental aspects in contemporary evolutionary theory. Parasitism is\u0000ubiquitous, inevitable, and arises as soon as the first replicators appear,\u0000even during the prebiotic stages of evolution. Both in theoretical approaches\u0000(computer modeling and analysis) and in real experiments (replication of\u0000biological macromolecules), parasitic processes emerge almost immediately. An\u0000effective way to avoid the elimination of the host-parasite system is through\u0000compartmentalization. In both theory and experiments, the pressure of\u0000parasitism leads to the complexification of the host-parasite system into a\u0000network of cooperative replicators and their parasites. Parasites have the\u0000ability to create niches for new replicators. The co-evolutionary arms race\u0000between defense systems and counter-defense mechanisms among parasites and\u0000hosts can progress for a considerable duration, involving multiple stages, if\u0000not indefinitely.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139063871","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Martin Falk, Adam Strupp, Benjamin Scellier, Arvind Murugan
Learning algorithms based on backpropagation have enabled transformative technological advances but alternatives based on local energy-based rules offer benefits in terms of biological plausibility and decentralized training. A broad class of such local learning rules involve textit{contrasting} a clamped configuration with the free, spontaneous behavior of the system. However, comparisons of clamped and free configurations require explicit memory or switching between Hebbian and anti-Hebbian modes. Here, we show how a simple form of implicit non-equilibrium memory in the update dynamics of each ``synapse'' of a network naturally allows for contrastive learning. During training, free and clamped behaviors are shown in sequence over time using a sawtooth-like temporal protocol that breaks the symmetry between those two behaviors when combined with non-equilibrium update dynamics at each synapse. We show that the needed dynamics is implicit in integral feedback control, broadening the range of physical and biological systems naturally capable of contrastive learning. Finally, we show that non-equilibrium dissipation improves learning quality and determine the Landauer energy cost of contrastive learning through physical dynamics.
{"title":"Contrastive learning through non-equilibrium memory","authors":"Martin Falk, Adam Strupp, Benjamin Scellier, Arvind Murugan","doi":"arxiv-2312.17723","DOIUrl":"https://doi.org/arxiv-2312.17723","url":null,"abstract":"Learning algorithms based on backpropagation have enabled transformative\u0000technological advances but alternatives based on local energy-based rules offer\u0000benefits in terms of biological plausibility and decentralized training. A\u0000broad class of such local learning rules involve textit{contrasting} a clamped\u0000configuration with the free, spontaneous behavior of the system. However,\u0000comparisons of clamped and free configurations require explicit memory or\u0000switching between Hebbian and anti-Hebbian modes. Here, we show how a simple\u0000form of implicit non-equilibrium memory in the update dynamics of each\u0000``synapse'' of a network naturally allows for contrastive learning. During\u0000training, free and clamped behaviors are shown in sequence over time using a\u0000sawtooth-like temporal protocol that breaks the symmetry between those two\u0000behaviors when combined with non-equilibrium update dynamics at each synapse.\u0000We show that the needed dynamics is implicit in integral feedback control,\u0000broadening the range of physical and biological systems naturally capable of\u0000contrastive learning. Finally, we show that non-equilibrium dissipation\u0000improves learning quality and determine the Landauer energy cost of contrastive\u0000learning through physical dynamics.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139063013","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tamas Veres, Mark Kerestely, Borbala M. Kovacs, David Keresztes, Klara Schulc, Erik Seitz, Zsolt Vassy, Daniel V. Veres, Peter Csermely
Recent findings show that single, non-neuronal cells are also able to learn signalling responses developing cellular memory. In cellular learning nodes of signalling networks strengthen their interactions e.g. by the conformational memory of intrinsically disordered proteins, protein translocation, miRNAs, lncRNAs, chromatin memory and signalling cascades. This can be described by a generalized, unicellular Hebbian learning process, where those signalling connections, which participate in learning, become stronger. Here we review those scenarios, where cellular signalling is not only repeated in a few times (when learning occurs), but becomes too frequent, too large, or too complex and overloads the cell. This leads to desensitisation of signalling networks by decoupling signalling components, receptor internalization, and consequent downregulation. These molecular processes are examples of anti-Hebbian learning and forgetting of signalling networks. Stress can be perceived as signalling overload inducing the desensitisation of signalling pathways. Aging occurs by the summative effects of cumulative stress downregulating signalling. We propose that cellular learning desensitisation, stress and aging may be placed along the same axis of more and more intensive (prolonged or repeated) signalling. We discuss how cells might discriminate between repeated and unexpected signals, and highlight the Hebbian and anti-Hebbian mechanisms behind the fold-change detection in the NF-k{appa}B signalling pathway. We list drug design methods using Hebbian learning (such as chemically-induced proximity) and clinical treatment modalities inducing (cancer, drug allergies) desensitisation or avoiding drug-induced desensitisation. A better discrimination between cellular learning, desensitisation and stress may open novel directions in drug design, e.g., helping to overcome drug-resistance.
{"title":"Cellular forgetting, desensitisation, stress and aging in signalling networks. When do cells refuse to learn more?","authors":"Tamas Veres, Mark Kerestely, Borbala M. Kovacs, David Keresztes, Klara Schulc, Erik Seitz, Zsolt Vassy, Daniel V. Veres, Peter Csermely","doi":"arxiv-2312.16875","DOIUrl":"https://doi.org/arxiv-2312.16875","url":null,"abstract":"Recent findings show that single, non-neuronal cells are also able to learn\u0000signalling responses developing cellular memory. In cellular learning nodes of\u0000signalling networks strengthen their interactions e.g. by the conformational\u0000memory of intrinsically disordered proteins, protein translocation, miRNAs,\u0000lncRNAs, chromatin memory and signalling cascades. This can be described by a\u0000generalized, unicellular Hebbian learning process, where those signalling\u0000connections, which participate in learning, become stronger. Here we review\u0000those scenarios, where cellular signalling is not only repeated in a few times\u0000(when learning occurs), but becomes too frequent, too large, or too complex and\u0000overloads the cell. This leads to desensitisation of signalling networks by\u0000decoupling signalling components, receptor internalization, and consequent\u0000downregulation. These molecular processes are examples of anti-Hebbian learning\u0000and forgetting of signalling networks. Stress can be perceived as signalling\u0000overload inducing the desensitisation of signalling pathways. Aging occurs by\u0000the summative effects of cumulative stress downregulating signalling. We\u0000propose that cellular learning desensitisation, stress and aging may be placed\u0000along the same axis of more and more intensive (prolonged or repeated)\u0000signalling. We discuss how cells might discriminate between repeated and\u0000unexpected signals, and highlight the Hebbian and anti-Hebbian mechanisms\u0000behind the fold-change detection in the NF-k{appa}B signalling pathway. We\u0000list drug design methods using Hebbian learning (such as chemically-induced\u0000proximity) and clinical treatment modalities inducing (cancer, drug allergies)\u0000desensitisation or avoiding drug-induced desensitisation. A better\u0000discrimination between cellular learning, desensitisation and stress may open\u0000novel directions in drug design, e.g., helping to overcome drug-resistance.","PeriodicalId":501325,"journal":{"name":"arXiv - QuanBio - Molecular Networks","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139062948","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：