RESUMO
Recent progress in deciphering mechanisms of human brain cortical folding leave unexplained whether spatially patterned genetic influences contribute to this folding. High-resolution in vivo brain MRI can be used to estimate genetic correlations (covariability due to shared genetic factors) in interregional cortical thickness, and biomechanical studies predict an influence of cortical thickness on folding patterns. However, progress has been hampered because shared genetic influences related to folding patterns likely operate at a scale that is much more local (<1 cm) than that addressed in prior imaging studies. Here, we develop methodological approaches to examine local genetic influences on cortical thickness and apply these methods to two large, independent samples. We find that such influences are markedly heterogeneous in strength, and in some cortical areas are notably stronger in specific orientations relative to gyri or sulci. The overall, phenotypic local correlation has a significant basis in shared genetic factors and is highly symmetric between left and right cortical hemispheres. Furthermore, the degree of local cortical folding relates systematically with the strength of local correlations, which tends to be higher in gyral crests and lower in sulcal fundi. The relationship between folding and local correlations is stronger in primary sensorimotor areas and weaker in association areas such as prefrontal cortex, consistent with reduced genetic constraints on the structural topology of association cortex. Collectively, our results suggest that patterned genetic influences on cortical thickness, measurable at the scale of in vivo MRI, may be a causal factor in the development of cortical folding.
Assuntos
Córtex Cerebral/anatomia & histologia , Córtex Cerebral/crescimento & desenvolvimento , Córtex Pré-Frontal/crescimento & desenvolvimento , Adulto , Idoso , Idoso de 80 Anos ou mais , Encéfalo/embriologia , Encéfalo/crescimento & desenvolvimento , Córtex Cerebral/metabolismo , Bases de Dados Factuais , Feminino , Humanos , Imageamento por Ressonância Magnética/métodos , Masculino , Pessoa de Meia-Idade , Córtex Pré-Frontal/anatomia & histologiaRESUMO
Neural systems are well known for their ability to learn and store information as memories. Even more impressive is their ability to abstract these memories to create complex internal representations, enabling advanced functions such as the spatial manipulation of mental representations. While recurrent neural networks (RNNs) are capable of representing complex information, the exact mechanisms of how dynamical neural systems perform abstraction are still not well-understood, thereby hindering the development of more advanced functions. Here, we train a 1000-neuron RNN-a reservoir computer (RC)-to abstract a continuous dynamical attractor memory from isolated examples of dynamical attractor memories. Furthermore, we explain the abstraction mechanism with a new theory. By training the RC on isolated and shifted examples of either stable limit cycles or chaotic Lorenz attractors, the RC learns a continuum of attractors as quantified by an extra Lyapunov exponent equal to zero. We propose a theoretical mechanism of this abstraction by combining ideas from differentiable generalized synchronization and feedback dynamics. Our results quantify abstraction in simple neural systems, enabling us to design artificial RNNs for abstraction and leading us toward a neural basis of abstraction.
Assuntos
Aprendizagem , Rede Nervosa , Computadores , Retroalimentação , Redes Neurais de ComputaçãoRESUMO
Regardless of the marked differences between biological and artificial neural systems, one fundamental similarity is that they are essentially dynamical systems that can learn to imitate other dynamical systems whose governing equations are unknown. The brain is able to learn the dynamic nature of the physical world via experience; analogously, artificial neural systems such as reservoir computing networks (RCNs) can learn the long-term behavior of complex dynamical systems from data. Recent work has shown that the mechanism of such learning in RCNs is invertible generalized synchronization (IGS). Yet, whether IGS is also the mechanism of learning in biological systems remains unclear. To shed light on this question, we draw inspiration from features of the human brain to propose a general and biologically feasible learning framework that utilizes IGS. To evaluate the framework's relevance, we construct several distinct neural network models as instantiations of the proposed framework. Regardless of their particularities, these neural network models can consistently learn to imitate other dynamical processes with a biologically feasible adaptation rule that modulates the strength of synapses. Further, we observe and theoretically explain the spontaneous emergence of four distinct phenomena reminiscent of cognitive functions: (i) learning multiple dynamics; (ii) switching among the imitations of multiple dynamical systems, either spontaneously or driven by external cues; (iii) filling-in missing variables from incomplete observations; and (iv) deciphering superimposed input from different dynamical systems. Collectively, our findings support the notion that biological neural networks can learn the dynamic nature of their environment through the mechanism of IGS.
Assuntos
Aprendizagem/fisiologia , Rede Nervosa/fisiologia , Redes Neurais de Computação , Humanos , Modelos Neurológicos , SinapsesRESUMO
Whether listening to overlapping conversations in a crowded room or recording the simultaneous electrical activity of millions of neurons, the natural world abounds with sparse measurements of complex overlapping signals that arise from dynamical processes. While tools that separate mixed signals into linear sources have proven necessary and useful, the underlying equational forms of most natural signals are unknown and nonlinear. Hence, there is a need for a framework that is general enough to extract sources without knowledge of their generating equations and flexible enough to accommodate nonlinear, even chaotic, sources. Here, we provide such a framework, where the sources are chaotic trajectories from independently evolving dynamical systems. We consider the mixture signal as the sum of two chaotic trajectories and propose a supervised learning scheme that extracts the chaotic trajectories from their mixture. Specifically, we recruit a complex dynamical system as an intermediate processor that is constantly driven by the mixture. We then obtain the separated chaotic trajectories based on this intermediate system by training the proper output functions. To demonstrate the generalizability of this framework in silico, we employ a tank of water as the intermediate system and show its success in separating two-part mixtures of various chaotic trajectories. Finally, we relate the underlying mechanism of this method to the state-observer problem. This relation provides a quantitative theory that explains the performance of our method, and why separation is difficult when two source signals are trajectories from the same chaotic system.
RESUMO
Cyclin-dependent kinase 8 (CDK8) has been identified as a colon cancer oncogene. Since this initial observation, CDK8 has been implicated as a potential driver of other cancers including acute myelogenous leukemia (AML) and some breast cancers. Here, we observed different biological responses to CDK8 inhibition among colon cancer cell lines and the triple-negative breast cancer (TNBC) cell line MDA-MB-468. When treated with CDK8 inhibitor 4, all treated cell lines responded with decreased cell viability and increased apoptosis. In the MDA-MB-468 cell line, the decrease in cell viability was dependent on increased phosphorylation of signal transducer and activator of transcription 3 (STAT3), which is not observed in the colon cancer cell lines. Furthermore, increased STAT3 phosphorylation in 4 treated MDA-MB-468 cells was dependent on increased transcription factor E2F1 protein. These results are consistent with previous reports of exogenous expression of E2F1-induced apoptosis in MDA-MB-468 cells.
Assuntos
Antineoplásicos/farmacologia , Apoptose/efeitos dos fármacos , Quinase 8 Dependente de Ciclina/metabolismo , Fator de Transcrição E2F1/metabolismo , Fosforilação/efeitos dos fármacos , Fator de Transcrição STAT3/metabolismo , Neoplasias de Mama Triplo Negativas/tratamento farmacológico , Neoplasias de Mama Triplo Negativas/metabolismo , Apoptose/fisiologia , Linhagem Celular Tumoral , Sobrevivência Celular/efeitos dos fármacos , Feminino , Células HCT116 , Humanos , Transdução de Sinais/efeitos dos fármacosRESUMO
We demonstrate the effectiveness of using machine learning for model-free prediction of spatiotemporally chaotic systems of arbitrarily large spatial extent and attractor dimension purely from observations of the system's past evolution. We present a parallel scheme with an example implementation based on the reservoir computing paradigm and demonstrate the scalability of our scheme using the Kuramoto-Sivashinsky equation as an example of a spatiotemporally chaotic system.
RESUMO
A machine-learning approach called "reservoir computing" has been used successfully for short-term prediction and attractor reconstruction of chaotic dynamical systems from time series data. We present a theoretical framework that describes conditions under which reservoir computing can create an empirical model capable of skillful short-term forecasts and accurate long-term ergodic behavior. We illustrate this theory through numerical experiments. We also argue that the theory applies to certain other machine learning methods for time series prediction.
RESUMO
We use recent advances in the machine learning area known as "reservoir computing" to formulate a method for model-free estimation from data of the Lyapunov exponents of a chaotic process. The technique uses a limited time series of measurements as input to a high-dimensional dynamical system called a "reservoir." After the reservoir's response to the data is recorded, linear regression is used to learn a large set of parameters, called the "output weights." The learned output weights are then used to form a modified autonomous reservoir designed to be capable of producing an arbitrarily long time series whose ergodic properties approximate those of the input signal. When successful, we say that the autonomous reservoir reproduces the attractor's "climate." Since the reservoir equations and output weights are known, we can compute the derivatives needed to determine the Lyapunov exponents of the autonomous reservoir, which we then use as estimates of the Lyapunov exponents for the original input generating system. We illustrate the effectiveness of our technique with two examples, the Lorenz system and the Kuramoto-Sivashinsky (KS) equation. In the case of the KS equation, we note that the high dimensional nature of the system and the large number of Lyapunov exponents yield a challenging test of our method, which we find the method successfully passes.
RESUMO
Deducing the state of a dynamical system as a function of time from a limited number of concurrent system state measurements is an important problem of great practical utility. A scheme that accomplishes this is called an "observer." We consider the case in which a model of the system is unavailable or insufficiently accurate, but "training" time series data of the desired state variables are available for a short period of time, and a limited number of other system variables are continually measured. We propose a solution to this problem using networks of neuron-like units known as "reservoir computers." The measurements that are continually available are input to the network, which is trained with the limited-time data to output estimates of the desired state variables. We demonstrate our method, which we call a "reservoir observer," using the Rössler system, the Lorenz system, and the spatiotemporally chaotic Kuramoto-Sivashinsky equation. Subject to the condition of observability (i.e., whether it is in principle possible, by any means, to infer the desired unmeasured variables from the measured variables), we show that the reservoir observer can be a very effective and versatile tool for robustly reconstructing unmeasured dynamical system variables.
RESUMO
Cells in the brain's Suprachiasmatic Nucleus (SCN) are known to regulate circadian rhythms in mammals. We model synchronization of SCN cells using the forced Kuramoto model, which consists of a large population of coupled phase oscillators (modeling individual SCN cells) with heterogeneous intrinsic frequencies and external periodic forcing. Here, the periodic forcing models diurnally varying external inputs such as sunrise, sunset, and alarm clocks. We reduce the dimensionality of the system using the ansatz of Ott and Antonsen and then study the effect of a sudden change of clock phase to simulate cross-time-zone travel. We estimate model parameters from previous biological experiments. By examining the phase space dynamics of the model, we study the mechanism leading to the difference typically experienced in the severity of jet-lag resulting from eastward and westward travel.
Assuntos
Ritmo Circadiano , Síndrome do Jet Lag/fisiopatologia , Viagem , Humanos , Núcleo Supraquiasmático/fisiopatologiaRESUMO
In gram-negative bacteria, the assembly of outer membrane proteins (OMPs) requires a ß-barrel assembly machinery (BAM) complex, of which BamA is an essential and evolutionarily conserved component. To elucidate the mechanism of BamA-mediated OMP biogenesis, we determined the crystal structure of the C-terminal transmembrane domain of BamA from Escherichia coli (EcBamA) at 2.6 Å resolution. The structure reveals 2 distinct features. First, a portion of the extracellular side of the ß barrel is composed of 5 markedly short ß strands, and the loops stemming from these ß strands form a potential surface cavity, filled by a portion of the L6 loop that includes the conserved VRGF/Y motif found in the Omp85 family. Second, the 4 extracellular loops L3, L4, L6, and L7 of EcBamA form a dome over the barrel, stabilized by a salt-bridge interaction network. Functional data show that hydrophilic-to-hydrophobic mutations of the potential hydrophilic surface cavity and a single Arg547Ala point mutation that may destabilize the dome severely affect the function of EcBamA. Our structure of the EcBamA ß barrel and structure-based mutagenesis studies suggest that the transmembrane ß strands of OMP substrates may integrate into the outer membrane at the interface of the first and last ß strands of the EcBamA barrel, whereas the soluble loops or domains may be transported out of the cell via the hydrophilic surface cavity on dislocation of the VRGF/Y motif of L6. In addition, the dome over the barrel may play an important role in maintaining the efficiency of OMP biogenesis.
Assuntos
Proteínas da Membrana Bacteriana Externa/química , Proteínas de Escherichia coli/química , Escherichia coli/química , Sequência de Aminoácidos , Proteínas da Membrana Bacteriana Externa/genética , Cristalografia por Raios X , Escherichia coli/genética , Proteínas de Escherichia coli/genética , Interações Hidrofóbicas e Hidrofílicas , Modelos Moleculares , Estrutura Terciária de ProteínaRESUMO
Accurate wheat ear counting is one of the key indicators for wheat phenotyping. Convolutional neural network (CNN) algorithms for counting wheat have evolved into sophisticated tools, however because of the limitations of sensory fields, CNN is unable to simulate global context information, which has an impact on counting performance. In this study, we present a hybrid attention network (CTHNet) for wheat ear counting from RGB images that combines local features and global context information. On the one hand, to extract multi-scale local features, a convolutional neural network is built using the Cross Stage Partial framework. On the other hand, to acquire better global context information, tokenized image patches from convolutional neural network feature maps are encoded as input sequences using Pyramid Pooling Transformer. Then, the feature fusion module merges the local features with the global context information to significantly enhance the feature representation. The Global Wheat Head Detection Dataset and Wheat Ear Detection Dataset are used to assess the proposed model. There were 3.40 and 5.21 average absolute errors, respectively. The performance of the proposed model was significantly better than previous studies.
RESUMO
In order to effectively support wheat breeding, farmland ridge segmentation can be used to visualize the size and spacing of a wheat field. At the same time, accurate ridge information collecting can deliver useful data support for farmland management. However, in the farming ridge segmentation scenarios based on remote sensing photos, the commonly used semantic segmentation methods tend to overlook the ridge edges and ridge strip features, which impair the segmentation effect. In order to efficiently collect ridge information, this paper proposes a segmentation method based on encoder-decoder of network with strip pooling module and ASPP module. First, in order to extract context information for multi-scale features, ASPP module are integrated in the deepest feature map. Second, the remote dependence of the ridge features is improved in both horizontal and vertical directions by using the strip pooling module. The final segmentation map is generated by fusing the boundary features and semantic features using an encoder and decoder architecture. As a result, the accuracy of the proposed method in the validation set is 98.0% and mIoU is 94.6%. The results of the experiments demonstrate that the method suggested in this paper can precisely segment the ridge information, as well as its value in obtaining data on the distribution of farmland and its potential for practical application.
RESUMO
RON is a receptor tyrosine kinase (RTK) of the MET receptor family that is canonically involved in mediating growth and inflammatory signaling. RON is expressed at low levels in a variety of tissues, but its overexpression and activation have been associated with malignancies in multiple tissue types and worse patient outcomes. RON and its ligand HGFL demonstrate cross-talk with other growth receptors and, consequentially, positions RON at the intersection of numerous tumorigenic signaling programs. For this reason, RON is an attractive therapeutic target in cancer research. A better understanding of homeostatic and oncogenic RON activity serves to enhance clinical insights in treating RON-expressing cancers.
Assuntos
Neoplasias , Proteínas Proto-Oncogênicas , Receptores Proteína Tirosina Quinases , Humanos , Fator de Crescimento de Hepatócito , Ligantes , Proteínas Proto-Oncogênicas/metabolismo , Transdução de SinaisRESUMO
In networks of coupled oscillators, it is of interest to understand how interaction topology affects synchronization. Many studies have gained key insights into this question by studying the classic Kuramoto oscillator model on static networks. However, new questions arise when the network structure is time varying or when the oscillator system is multistable, the latter of which can occur when an inertial term is added to the Kuramoto model. While the consequences of evolving topology and multistability on collective behavior have been examined separately, real-world systems such as gene regulatory networks and the brain may exhibit these properties simultaneously. It is thus relevant to ask how time-varying network connectivity impacts synchronization in systems that can exhibit multistability. To address this question, we study how the dynamics of coupled Kuramoto oscillators with inertia are affected when the topology of the underlying network changes in time. We show that hysteretic synchronization behavior in networks of coupled inertial oscillators can be driven by changes in connection topology alone. Moreover, we find that certain fixed-density rewiring schemes induce significant changes to the level of global synchrony that remain even after the network returns to its initial configuration, and we show that these changes are robust to a wide range of network perturbations. Our findings highlight that the specific progression of network topology over time, in addition to its initial or final static structure, can play a considerable role in modulating the collective behavior of systems evolving on complex networks.
RESUMO
Network control theory is increasingly used to profile the brain's energy landscape via simulations of neural dynamics. This approach estimates the control energy required to simulate the activation of brain circuits based on structural connectome measured using diffusion magnetic resonance imaging, thereby quantifying those circuits' energetic efficiency. The biological basis of control energy, however, remains unknown, hampering its further application. To fill this gap, investigating temporal lobe epilepsy as a lesion model, we show that patients require higher control energy to activate the limbic network than healthy volunteers, especially ipsilateral to the seizure focus. The energetic imbalance between ipsilateral and contralateral temporolimbic regions is tracked by asymmetric patterns of glucose metabolism measured using positron emission tomography, which, in turn, may be selectively explained by asymmetric gray matter loss as evidenced in the hippocampus. Our investigation provides the first theoretical framework unifying gray matter integrity, metabolism, and energetic generation of neural dynamics.
RESUMO
A mechanism, which is distinct from the traditional one when sodium alkoxide was used instead of tertiary amines, was proposed for the alkoxycarbonylation of aryl iodides. The catalytic cycle was composed of oxidative addition, subsequent ArPdOR formation, CO insertion to Pd-OR, and final reductive elimination of ArPdCOOR. The kinetic simultaneity of the formation of deiodinated side product from the aryl iodide and aldehyde from corresponding alcohol provided strong evidence for the existence of ArPdOR species. The observation of thioether, as the other competitive product in palladium catalyzed thiocarbonylation of aryl iodides and sodium alkylthiolate, also indicate the possibility of metathesis between ArPdI and sodium alkylthiolate. Preliminary kinetic studies revealed that neither oxidative addition nor reductive elimination was rate limiting. DFT calculation displayed preference for CO insertion into Pd-OR bond. The advantage of this novel mechanism had been demonstrated in the facile alkoxycarbonylation and thiocarbonylation. The ethoxycarbonylation of aryl iodides under room temperature and balloon pressure of CO in the presence of EtONa were examined, and good to high yields were obtained; the t-butoxycarbonylation reactions in the presence of t-BuONa were achieved, and the alkylthiocarbonylation (including the t-butylthiocarbonylation) of aryl iodides in the presence of sodium alkylthiolate were also investigated.
Assuntos
Hidrocarbonetos Iodados/síntese química , Compostos Organometálicos/química , Paládio/química , Álcoois/química , Monóxido de Carbono/química , Catálise , Simulação por Computador , Hidrocarbonetos Iodados/química , Estrutura Molecular , Oxirredução , EstereoisomerismoRESUMO
Executive function develops during adolescence, yet it remains unknown how structural brain networks mature to facilitate activation of the fronto-parietal system, which is critical for executive function. In a sample of 946 human youths (ages 8-23y) who completed diffusion imaging, we capitalized upon recent advances in linear dynamical network control theory to calculate the energetic cost necessary to activate the fronto-parietal system through the control of multiple brain regions given existing structural network topology. We found that the energy required to activate the fronto-parietal system declined with development, and the pattern of regional energetic cost predicts unseen individuals' brain maturity. Finally, energetic requirements of the cingulate cortex were negatively correlated with executive performance, and partially mediated the development of executive performance with age. Our results reveal a mechanism by which structural networks develop during adolescence to reduce the theoretical energetic costs of transitions to activation states necessary for executive function.
Adolescents are known for taking risks, from driving too fast to experimenting with drugs and alcohol. Such behaviors tend to decrease as individuals move into adulthood. Most people in their mid-twenties have greater self-control than they did as teenagers. They are also often better at planning, sustaining attention, and inhibiting impulsive behaviors. These skills, which are known as executive functions, develop over the course of adolescence. Executive functions rely upon a series of brain regions distributed across the frontal lobe and the lobe that sits just behind it, the parietal lobe. Fiber tracts connect these regions to form a fronto-parietal network. These fiber tracts are also referred to as white matter due to the whitish fatty material that surrounds and insulates them. Cui et al. now show that changes in white matter networks have implications for teen behavior. Almost 950 healthy young people aged between 8 and 23 years underwent a type of brain scan called diffusion-weighted imaging that visualizes white matter. The scans revealed that white matter networks in the frontal and parietal lobes mature over adolescence. This makes it easier for individuals to activate their fronto-parietal networks by decreasing the amount of energy required. Cui et al. show that a computer model can predict the maturity of a person's brain based on the energy needed to activate their fronto-parietal networks. These changes help explain why executive functions improve during adolescence. This in turn explains why behaviors such as risk-taking tend to decrease with age. That said, adults with various psychiatric disorders, such as ADHD and psychosis, often show impaired executive functions. In the future, it may be possible to reduce these impairments by applying magnetic fields to the scalp to reduce the activity of specific brain regions. The techniques used in the current study could help reveal which brain regions to target with this approach.
Assuntos
Mapeamento Encefálico , Encéfalo/fisiologia , Função Executiva/fisiologia , Vias Neurais/fisiologia , Adolescente , Mapeamento Encefálico/métodos , Criança , Imagem de Difusão por Ressonância Magnética/métodos , Feminino , Humanos , Imageamento por Ressonância Magnética/métodos , Masculino , Adulto JovemRESUMO
We study the firing dynamics of a discrete-state and discrete-time version of an integrate-and-fire neuronal network model with both excitatory and inhibitory neurons. When the integer-valued state of a neuron exceeds a threshold value, the neuron fires, sends out state-changing signals to its connected neurons, and returns to the resting state. In this model, a continuous phase transition from non-ceaseless firing to ceaseless firing is observed. At criticality, power-law distributions of avalanche size and duration with the previously derived exponents, -3/2 and -2, respectively, are observed. Using a mean-field approach, we show analytically how the critical point depends on model parameters. Our main result is that the combined presence of both inhibitory neurons and integrate-and-fire dynamics greatly enhances the robustness of critical power-law behavior (i.e., there is an increased range of parameters, including both sub- and supercritical values, for which several decades of power-law behavior occurs).
Assuntos
Modelos Neurológicos , Neurônios/fisiologia , Animais , Humanos , Rede Nervosa/fisiologiaRESUMO
In this paper we consider the motion of point particles in a particular type of one-degree-of-freedom, slowly changing, temporally periodic Hamiltonian. Through most of the time cycle, the particles conserve their action, but when a separatrix is approached and crossed, the conservation of action breaks down, as shown in previous theoretical studies. These crossings have the effect that the numerical solution shows an apparent contradiction. Specifically we consider two initial constant energy phase space curves H=E(A) and H=E(B) at time t=0, where H is the Hamiltonian and E(A) and E(B) are the two initial energies. The curve H=E(A) encircles the curve H=E(B). We then sprinkle many initial conditions (particles) on these curves and numerically follow their orbits from t=0 forward in time by one cycle period. At the end of the cycle the vast majority of points initially on the curves H=E(A) and H=E(B) now appear to lie on two new constant energy curves H=E(A)' and H=E(B)', where the B' curve now encircles the A' curve (as opposed to the initial case where the A curve encircles the B curve). Due to the uniqueness of Hamilton dynamics, curves evolved under the dynamics cannot cross each other. Thus the apparent curves H=E(A)' and H=E(B)' must be only approximate representations of the true situation that respects the topological exclusion of curve crossing. In this paper we resolve this apparent paradox and study its consequences. For this purpose we introduce a "robust" numerical simulation technique for studying the complex time evolution of a phase space curve in a Hamiltonian system. We also consider how a very tiny amount of friction can have a major consequence, as well as what happens when a very large number of cycles is followed. We also discuss how this phenomenon might extend to chaotic motion in higher dimensional Hamiltonian systems.