Gating Mechanism Research Articles

Zero-shot stance detection based on multi-expert collaboration

Zero-shot stance detection is pivotal for autonomously discerning user stances on novel emerging topics. This task hinges on effective feature alignment transfer from known to unseen targets. To address this, we introduce a zero-shot stance detection framework utilizing multi-expert cooperative learning. This framework comprises two core components: a multi-expert feature extraction module and a gating mechanism for stance feature selection. Our approach involves a unique learning strategy tailored to decompose complex semantic features. This strategy harnesses the expertise of multiple specialists to unravel and learn diverse, intrinsic textual features, enhancing transferability. Furthermore, we employ a gating-based mechanism to selectively filter and fuse these intricate features, optimizing them for stance classification. Extensive experiments on standard benchmark datasets demonstrate that our model significantly surpasses existing baseline models in performance.

Estimating the Health State of Lithium-Ion Batteries Using an Adaptive Gated Sequence Network and Hierarchical Feature Construction

State of health (SOH) estimation plays a vital role in ensuring the safe and stable operation of lithium-ion battery management systems (BMSs). Data-driven methods are widely used to estimate SOH; however, existing methods often suffer from fixed or excessively high feature dimensions, impacting the model’s adjustability and applicability. This study first proposed a layered knee point strategy based on the charging voltage curve, which reduced the complexity of feature extraction. Then, a new hybrid framework called the adaptive gated sequence network (AGSN) model was proposed. This model integrated independently recurrent neural network (IndRNN) layers, active state tracking long short-term memory (AST-LSTM) layers, and adaptive gating mechanism (AGM) layers. By integrating a multi-layered structure and an adaptive gating mechanism, the SOH prediction performance was significantly improved. Finally, batteries under different operating conditions were tested using the NASA battery dataset. The results show that the AGSN model demonstrated higher accuracy and robustness in battery SOH estimation, with estimation errors consistently within 1%.

Effect of two activators on the gating of a K2P channel

TWIK-related potassium channel 1 (TREK1), a two-pore-domain mammalian potassium (K+) channel, regulates the resting potential across cell membranes, presenting a promising therapeutic target for neuropathy treatment. The gating of this channel converges in the conformation of the narrowest part of the pore: the selectivity filter (SF). Various hypotheses explain TREK1 gating modulation, including the dynamics of loops connecting the SF with transmembrane helices and the stability of hydrogen bond (HB) networks adjacent to the SF. Recently, two small molecules (Q6F and Q5F) were reported as activators that affect TREK1 by increasing its open probability in single-channel current measurements. Here, using molecular dynamics simulations, we investigate the effect of these ligands on the previously proposed modulation mechanisms of TREK1 gating compared to the apo channel. Our findings reveal that loop dynamics at the upper region of the SF exhibit only a weak correlation with permeation events/nonpermeation periods, whereas the HB network behind the SF appears more correlated. These nonpermeation periods arise from both distinct mechanisms: a C-type inactivation (resulting from dilation at the top of the SF), which has been described previously, and a carbonyl flipping in an SF binding site. We find that, besides the prevention of C-type inactivation in the channel, the ligands increase the probability of permeation by modulating the dynamics of the carbonyl flipping, influenced by a threonine residue at the bottom of the SF. These results offer insights for rational ligand design to optimize the gating modulation of TREK1 and related K+ channels.

Molecular mechanism of ligand gating and opening of NMDA receptor.

Glutamate transmission and activation of ionotropic glutamate receptors are the fundamental means by which neurons control their excitability and neuroplasticity1. The N-methyl-D-aspartate receptor (NMDAR) is unique among all ligand-gated channels, requiring two ligands-glutamate and glycine-for activation. These receptors function as heterotetrameric ion channels, with the channel opening dependent on the simultaneous binding of glycine and glutamate to the extracellular ligand-binding domains (LBDs) of the GluN1 and GluN2 subunits, respectively2,3. The exact molecular mechanism for channel gating by the two ligands has been unclear, particularly without structures representing the open channel and apo states. Here we show that the channel gate opening requires tension in the linker connecting the LBD and transmembrane domain (TMD) and rotation of the extracellular domain relative to the TMD. Using electron cryomicroscopy, we captured the structure of theGluN1-GluN2B(GluN1-2B) NMDAR in its open state bound to a positive allosteric modulator. This process rotates and bends the pore-forming helices in GluN1 and GluN2B, altering the symmetry of the TMD channel from pseudofourfold to twofold. Structures of GluN1-2B NMDAR in apo and single-liganded states showed that binding of either glycine or glutamate alone leads to distinct GluN1-2B dimer arrangements but insufficient tension in the LBD-TMD linker for channel opening. This mechanistic framework identifies a key determinant for channel gating and a potential pharmacological strategy for modulating NMDAR activity.

Enhancing stock market prediction accuracy with recurrent deep learning models: A case study on the CAC40 index

This paper explores the application of Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM) networks for stock price prediction over a 10-day horizon. The study aims to compare the predictive performance of these two deep learning architectures within the context of financial forecasting. Utilizing historical stock data from the CAC40 dataset, which represents a capitalization-weighted measure of the 40 most significant stocks on the Euronext Paris, we train and evaluate RNN and LSTM models to forecast future stock prices. Our results demonstrate the superior performance of LSTM networks in capturing the intricate temporal dependencies inherent in stock price data. Compared to standard RNNs, LSTM models exhibit higher accuracy and provide more reliable forecasts over the 10-day prediction period. The specialized memory cells and gating mechanisms in LSTM networks enable them to effectively identify both short-term changes and long-term patterns in stock prices, thus outperforming traditional RNN architectures. This enhanced ability to model the complex dynamics of stock market data underscores the potential of LSTM networks to improve investment decision-making, risk management, and the overall efficiency of financial markets. The insights gained from this study contribute to the growing body of knowledge on the application of deep learning in finance and investment, offering valuable guidance for practitioners and researchers seeking to harness the power of advanced algorithms for stock market prediction and optimization.

Periphery Pre-S1 and S1 helix nexus for PIP2 at TRPC3 channel

Transient receptor potential canonical 3 (TRPC3) is a calcium-permeable, non-selective cation channel known to be regulated by components of the phospholipase C (PLC)-mediated signaling pathway, such as Ca2+, diacylglycerol (DAG) and phosphatidylinositol 4,5-biphosphate (PI(4,5)P2). However, the molecular gating mechanism by these regulators is not yet fully understood, especially its regulation by PI(4,5)P2, despite the importance of this channel in cardiovascular pathophysiology. Recently, Clarke et al. (2024) have reported that PI(4,5)P2 is a positive modulator for TRPC3 using molecular dynamics simulations and patch-clamp techniques. They have demonstrated a multistep gating mechanism of TRPC3 with the binding of PI(4,5)P2 to the lipid binding site located at the pre-S1/S1 nexus, and the propagation of PI(4,5)P2 sensing to the pore domain via a salt bridge between the TRP helix and the S4–S5 linker.

Structured flexibility in recurrent neural networks via neuromodulation.

The goal of theoretical neuroscience is to develop models that help us better understand biological intelligence. Such models range broadly in complexity and biological detail. For example, task-optimized recurrent neural networks (RNNs) have generated hypotheses about how the brain may perform various computations, but these models typically assume a fixed weight matrix representing the synaptic connectivity between neurons. From decades of neuroscience research, we know that synaptic weights are constantly changing, controlled in part by chemicals such as neuromodulators. In this work we explore the computational implications of synaptic gain scaling, a form of neuromodulation, using task-optimized low-rank RNNs. In our neuromodulated RNN (NM-RNN) model, a neuromodulatory subnetwork outputs a low-dimensional neuromodulatory signal that dynamically scales the low-rank recurrent weights of an output-generating RNN. In empirical experiments, we find that the structured flexibility in the NM-RNN allows it to both train and generalize with a higher degree of accuracy than low-rank RNNs on a set of canonical tasks. Additionally, via theoretical analyses we show how neuromodulatory gain scaling endows networks with gating mechanisms commonly found in artificial RNNs. We end by analyzing the low-rank dynamics of trained NM-RNNs, to show how task computations are distributed.

Research on time-series financial data prediction and analysis based on deep recurrent neural network

Time series data is widely available in a variety of industries. By forecasting time series, decision-makers can better grasp future trends and make more effective decisions. Financial time series data exhibit non-stationarity and high volatility. High-frequency fluctuations in financial products such as exchange rates, bonds and equities may reflect external shocks and risks in global financial markets, which are potentially dangerous and may threaten national economic security or even trigger financial crises. For financial time series data, a deep recurrent neural network first progressively processes each data point in the time series through its recurrent unit. Each recurring unit can adjust its own weights to better predict or analyze future values. Over time, these recurrent units continuously update their internal state, resulting in a comprehensive understanding of the characteristics of the entire data sequence. In addition, we add a gating mechanism to further improve the network's ability to control the flow of information, so that the model is more effective when retaining long-term dependencies, so as to improve the accuracy of prediction and the stability of the model. Experimental results show that our recurrent neural network model shows higher prediction accuracy and stability than other baseline models on financial time series datasets.

ICKA: An instruction construction and Knowledge Alignment framework for Multimodal Named Entity Recognition

Multimodal Named Entity Recognition (MNER) aims to identify entities of predefined types in text by leveraging information from multiple modalities, most notably textual and visual information. Most efforts concentrate on improving cross-modality attention mechanisms to facilitate guidance between modalities. However, they still suffer from certain limitations: (1) it is difficult to establish a unified representation to bridge the semantic gap among different modalities; (2) mining the implicit relationships between text and image is crucial yet challenging. In this paper, we propose an Instruction Construction and Knowledge Alignment Framework for MNER named ICKA to address these issues. Specifically, we first employ a multi-head cross-modal attention mechanism to obtain the cross-modal fusion representation by fusing features from text–image pairs. Then, we integrate external knowledge from the pre-trained vision-language model (VLM) to facilitate semantic alignment between text and image and obtain inter-modality connections. Next, we construct the multimodal instruction that consists of the modal features and uses the inter-modality connections as a bridge between them. We then integrate the instruction into the language model to effectively incorporate multimodal knowledge. Finally, we perform sequence labeling using a Conditional Random Fields (CRF) decoder with a gating mechanism. The proposed method achieves F1 scores of 75.42% on the Twitter2015 dataset and 87.12% on the Twitter2017 dataset, demonstrating the competitiveness of our method.

Telomerase Reverse Transcriptase Regulates Intracellular Ca2+ Homeostasis and Mitochondrial Function via the p53/PGC-1α Pathway in HL-1 Cells.

Telomere shortening is strongly associated with cardiovascular aging and disease, and patients with shorter telomeres in peripheral blood leukocytes are at higher risk of cardiovascular diseases such as heart failure and atrial fibrillation (AF). Telomerase reverse transcriptase (TERT) maintains telomere length, and overexpression of TERT has been shown to reduce cardiomyocyte apoptosis and myocardial infarct size, and extend the lifespan of aged mice. However, the specific impact of TERT on the electrophysiology of cardiomyocytes remains to be elucidated. The aims of this study were to evaluate the role of TERT in Ca2+ homeostasis and mitochondrial function in atrial myocytes as well as the underlying mechanisms. TERT overexpressed and silenced HL-1 cells were constructed with lentiviruses, and the respective empty lentiviral vectors were used as negative controls. Then the patch clamp technique was used to record the electrophysiological characteristics such as cell action potential duration (APD) and L-type Ca2+ currents (ICa,L), flow cytometry was used to detect intracellular Ca2+ concentration and mitochondrial membrane potential (MMP), and the Seahorse assay was used to measure the oxygen consumption rate (OCR). TERT silencing led to intracellular Ca2+ overload, shortened APD, decreased ICa,L current density, altered Ca2+ gating mechanism, decreased MMP and OCR, and increased reactive oxygen species (ROS), whereas TERT overexpression led to the reverse effects. Additionally, TERT silencing resulted in intracellular Ca2+ overload with decreased expression of the SERCA2a, CaV1.2, and NCX1.1, whereas TERT overexpression had opposing effects. Furthermore, we discovered that TERT could regulate the expression of p53 and peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α). The expression of PGC-1α was downregulated by the p53 agonist Tenovin-6 but upregulated by the p53 inhibitor PFTα. The effects of the PGC-1α inhibitor SR-18292 on intracellular Ca2+ and cell electrophysiology were similar to those of silencing TERT, whereas the PGC-1α agonist ZLN005 produced comparable outcomes to TERT overexpression. TERT silencing-induced Ca2+ overload and mitochondrial dysfunction may be one mechanism of age-related AF. Overexpression of TERT reduced the basis for arrhythmia formation such as AF, suggesting a favorable safety profile for TERT therapy. TERT regulated intracellular Ca2+ homeostasis and mitochondrial function through the p53/PGC-1α pathway. In addition, PGC-1α might be a novel target for AF, suggesting that intervention for AF should be not limited to abnormal cation handling.

Guided-attention and gated-aggregation network for medical image segmentation

Recently, transformers have been widely used in medical image segmentation to capture long-range and global dependencies using self-attention. However, they often struggle to learn the local details which limit their ability to capture irregular shapes and sizes of the tissues and indistinct boundaries between the tissues, which are critical for accurate segmentation. To alleviate this issue, we propose a network named GA2Net, which comprises an encoder, a bottleneck, and a decoder. The encoder computes multi-scale features. In the bottleneck, we propose a hierarchical-gated features aggregation (HGFA) which introduces a novel spatial gating mechanism to enrich the multi-scale features. To effectively learn the shapes and sizes of the tissues, we apply deep supervision in the bottleneck. GA2Net proposes to use adaptive aggregation (AA) within the decoder, to adjust the receptive fields for each location in the feature map, by replacing the traditional concatenation/summation operations in skip connections in U-Net like architecture. Furthermore, we propose mask-guided feature attention (MGFA) modules within the decoder which strives to learn the salient features using foreground priors to adequately grasp the intricate structural and contour information of the tissues. We also apply intermediate supervision for each stage of the decoder, which further improves the capability of the model to better locate the boundaries of the tissues. Our extensive experimental results illustrate that our GA2-Net significantly outperforms the existing state-of-the-art methods over eight medical image segmentation datasets i.e., five polyps, a skin lesion, a multiple myeloma cell segmentation, and a cardiac MRI scan datasets. We then perform an extensive ablation study to validate the capabilities of our method. Code is available at https://github.com/mustansarfiaz/ga2net.

Pre-gating and contextual attention gate — A new fusion method for multi-modal data tasks

Multi-modal representation learning has received significant attention across diverse research domains due to its ability to model a scenario comprehensively. Learning the cross-modal interactions is essential to combining multi-modal data into a joint representation. However, conventional cross-attention mechanisms can produce noisy and non-meaningful values in the absence of useful cross-modal interactions among input features, thereby introducing uncertainty into the feature representation. These factors have the potential to degrade the performance of downstream tasks. This paper introduces a novel Pre-gating and Contextual Attention Gate (PCAG) module for multi-modal learning comprising two gating mechanisms that operate at distinct information processing levels within the deep learning model. The first gate filters out interactions that lack informativeness for the downstream task, while the second gate reduces the uncertainty introduced by the cross-attention module. Experimental results on eight multi-modal classification tasks spanning various domains show that the multi-modal fusion model with PCAG outperforms state-of-the-art multi-modal fusion models. Additionally, we elucidate how PCAG effectively processes cross-modality interactions

Enhanced Air Quality Prediction through Spatio-temporal Feature Sxtraction and Fusion: A Self-tuning Hybrid Approach with GCN and GRU

Accurate prediction of air quality change is essential for air pollution control and human daily mobility. Due to the strong spatial and temporal correlation of air quality changes, existing air quality prediction methods often face the problem of low prediction accuracy due to insufficient extraction of spatio-temporal features. In this paper, we proposed a self-tuning spatio-temporal neural network (ST2NN) to enhance air quality prediction. ST2NN model consisted of four modules. First, ST2NN model constructed a temporal feature extraction module and a spatial feature extraction module based on gated recurrent unit (GRU) and graph convolutional neural network (GCN), respectively, and the two feature extraction modules adopted a parallel structure, which could effectively extract the spatio-temporal features in data. Additionally, ST2NN model constructed a feature fusion module based on gating mechanism to delineate the contribution of spatio-temporal features to the predicted values. Further, ST2NN model constructed a Hyperband hyperparameter optimization module based on Hyperband optimization algorithm to automatically adjust the network hyperparameters. The structure of ST2NN model endowed it with excellent spatio-temporal feature extraction and parameter adaptability. ST2NN model was evaluated and compared with existing models, including convolutional long short-term memory neural network (ConvLSTM), GRU, combined convolutional neural network and long short-term memory neural network (CNN-LSTM), and GCN-LSTM for air quality index (AQI) prediction using data from twelve monitoring stations in Beijing, China. Across all four evaluation indexes, ST2NN model outperformed the comparative models, improving prediction accuracy by 0.51%-10.18% (measured using R2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${R}^{2}$$\\end{document}). From the experimental results, it can be seen that ST2NN model constructed from the perspective of spatio-temporal feature extraction has better prediction performance compared with the existing air quality prediction model, which provides a new method for air quality prediction and has certain application value.

Insights into CLC-0's Slow-Gating from Intracellular Proton Inhibition.

The opening of the Torpedo CLC-0 chloride (Cl-) channel is known to be regulated by two gating mechanisms: fast gating and slow (common) gating. The structural basis underlying the fast-gating mechanism is better understood than that of the slow-gating mechanism, which is still largely a mystery. Our previous study on the intracellular proton (H+i)-induced inhibition of the CLC-0 anionic current led to the conclusion that the inhibition results from the slow-gate closure (also called inactivation). The conclusion was made based on substantial evidence such as a large temperature dependence of the H+i inhibition similar to that of the channel inactivation, a resistance to the H+i inhibition in the inactivation-suppressed C212S mutant, and a similar voltage dependence between the current recovery from the H+i inhibition and the recovery from the channel inactivation. In this work, we further examine the mechanism of the H+i inhibition of wild-type CLC-0 and several mutants. We observe that an anion efflux through the pore of CLC-0 accelerates the recovery from the H+i-induced inhibition, a process corresponding to the slow-gate opening. Furthermore, various inactivation-suppressed mutants exhibit different current recovery kinetics, suggesting the existence of multiple inactivated states (namely, slow-gate closed states). We speculate that protonation of the pore of CLC-0 increases the binding affinity of permeant anions in the pore, thereby generating a pore blockage of ion flow as the first step of inactivation. Subsequent complex protein conformational changes further transition the CLC-0 channel to deeper inactivated states.

An obstacle avoidance method for robotic arm based on reinforcement learning

Purpose Robotic arms play a crucial role in various industrial operations, such as sorting, assembly, handling and spraying. However, traditional robotic arm control algorithms often struggle to adapt when faced with the challenge of dynamic obstacles. This paper aims to propose a dynamic obstacle avoidance method based on reinforcement learning to address real-time processing of dynamic obstacles. Design/methodology/approach This paper introduces an innovative method that introduces a feature extraction network that integrates gating mechanisms on the basis of traditional reinforcement learning algorithms. Additionally, an adaptive dynamic reward mechanism is designed to optimize the obstacle avoidance strategy. Findings Validation through the CoppeliaSim simulation environment and on-site testing has demonstrated the method's capability to effectively evade randomly moving obstacles, with a significant improvement in the convergence speed compared to traditional algorithms. Originality/value The proposed dynamic obstacle avoidance method based on Reinforcement Learning not only accomplishes the task of dynamic obstacle avoidance efficiently but also offers a distinct advantage in terms of convergence speed. This approach provides a novel solution to the obstacle avoidance methods for robotic arms.

Electronic Polarization Leads to a Drier Dewetted State for Hydrophobic Gating in the Big Potassium Channel.

In the hydrophobic gating mechanism proposed for some ion channels, ion permeation is not blocked by the physical dimension of the channel pore but by its dewetted state which constitutes the energetic bottleneck. A major source of uncertainty in the mechanism is that the dewetted state was not observed in experiments and only probed in simulations using nonpolarizable force fields, which do not accurately represent the properties of confined water. Here we analyze hydration of the central cavity in the pore-gate domain of the Big Potassium channel using molecular dynamics and grand canonical Monte Carlo simulations with enhanced sampling techniques. Including polarization leads to a much drier dewetted state and a higher barrier for the transition to the wet state, suggesting more effective hydrophobic gating. The simulations also identify two backbone carbonyls at the bottom of the selectivity filter as good candidates for characterizing the dewetted state using infrared spectroscopies.

A deep neural network for operator learning enhanced by attention and gating mechanisms for long-time forecasting of tumor growth

A deep neural network for operator learning enhanced by attention and gating mechanisms for long-time forecasting of tumor growth

Novel behavior-enhanced long- and short-term interest model for sequential recommendation

In the realm of modern recommender systems, user-item interaction data often exhibit sequential patterns in relation to various behaviors, such as clicks and purchases on e-commerce platforms. The objective of heterogeneous sequential recommendation (HSR) is to predict each user's next item of interest under a specific behavior based on these interactions. However, existing approaches to HSR struggle to fully capture item transition relationships, as they consider these relationships from a single dimension and at a coarse-grained level. To bridge this gap, we propose the novel Behavior-enhanced Long- and Short-term Interest (BLSI) model, which explores fine-grained item transition relationships in both local and global dimensionalities. At its core, BLSI incorporates a behavior-enhanced self-attention network (BSAN) to capture short-term user preferences. BSAN distinguishes the effects of different behaviors and considers cross-type behavior influences during the linear projection and attention score calculation stages. Additionally, BLSI employs a heterogeneous graph neural network (HGNN) to model long-term user interests by discriminatively aggregating the information of neighboring nodes according to their behavior transition relationships. Furthermore, a gating mechanism is implemented to adaptively fuse short- and long-term preferences for personalized recommendations. Extensive experimental results on three datasets demonstrate that BLSI significantly outperforms state-of-the-art recommendation methods, highlighting the advantages of leveraging sequentiality and behavioral heterogeneity.

Interior pH-sensing residue of human voltage-gated proton channel Hv1 is histidine 168

The molecular mechanisms governing the human voltage-gated proton channel hHv1 remain elusive. Here, we used membrane-enabled hybrid-solvent continuous constant pH molecular dynamics (CpHMD) simulations with pH replica exchange to further evaluate the structural models of hHv1 in the closed (hyperpolarized) and open (depolarized) states recently obtained with MD simulations and explore potential pH-sensing residues. The CpHMD titration at a set of symmetric pH conditions revealed three residues that can gain or lose protons upon channel depolarization. Among them, residue H168 at the intracellular end of the S3 helix switches from the deprotonated to the protonated state and its protonation is correlated with the increased tilting of the S3 helix during the transition from the closed to the open state. Thus, the simulation data suggest H168 as an interior pH sensor, in support of a recent finding based on electrophysiological experiments of Hv1 mutants. We propose that protonation of H168 acts as a key that unlocks the closed channel configuration by increasing the flexibility of the S2-S3 linker, which increases the tilt angle of S3 and enhances the mobility of the S4 helix, thus promoting channel opening. Our work represents an important step toward deciphering the pH-dependent gating mechanism of hHv1.

Toxic Peptides from the Mexican Scorpion Centruroides villegasi: Chemical Structure and Evaluation of Recognition by Human Single-Chain Antibodies.

Alternative recombinant sources of antivenoms have been successfully generated. The application of such strategies requires the characterization of the venoms for the development of specific neutralizing molecules against the toxic components. Five toxic peptides to mammals from the Mexican scorpion Centruroides villegasi were isolated by chromatographic procedures by means of gel filtration on Sephadex G-50, followed by ion-exchange columns on carboxy-methyl-cellulose (CMC) resins and finally purified by high-performance chromatography (HPLC) columns. Their primary structures were determined by Edman degradation. They contain 66 amino acids and are maintained well packed by four disulfide bridges, with molecular mass from 7511.3 to 7750.1 Da. They are all relatively toxic and deadly to mice and show high sequence identity with known peptides that are specific modifiers of the gating mechanisms of Na+ ion channels of type beta-toxin (β-ScTx). They were named Cv1 to Cv5 and used to test their recognition by single-chain variable fragments (scFv) of antibodies, using surface plasmon resonance. Three different scFvs generated in our laboratory (10FG2, HV, LR) were tested for recognizing the various new peptides described here, paving the way for the development of a novel type of scorpion antivenom.