Propagation Network Research Articles

Network propagation unveils molecular mechanisms modulated in Alzheimer’s Disease patients treated by the complex plasma fraction GRF6019

AbstractBackgroundOur pre‐clinical work provided evidence of the interest of plasma fraction treatment to reverse age‐related cognitive decline. One of the most potent plasma fractions, GRF6019 was tested on patients with Alzheimer’s disease (AD). GRF6019 was safe and well tolerated, and patients experienced no cognitive decline and minimal functional decline. However, understanding these responses requires large‐scale mapping of the underlying molecular mechanisms. Recent affinity‐based approaches can measure levels of thousands of proteins in plasma but the increased signal‐to‐noise ratio and the large‐scale analysis of the proteomics changes require sophisticated bioinformatic techniques. To this end, we applied network propagation methods and combined protein‐protein interaction (PPI) networks, known AD‐associated proteins and GRF6019 induced protein changes to shed the light on the molecular mechanisms underlying patients’ responses to GRF6019 treatment.MethodPhase 2 clinical trials with repeated infusions of the plasma fraction GRF6019 were conducted in patients with AD (GRF6019‐201 n = 40 and GRF6019‐202 n = 26). Proteomic measurements from 136 plasma samples across three timepoints were performed by SomaLogic and Olink platforms. GRF6019‐induced proteomics changes were expressed by a Modulation Score (MS) for each protein. This MS captures the amplitude of the proteomics changes and the consistency of these changes within and between trials. Network propagation modeled the spread of MS on the PPI network. The network propagation algorithm we applied was the fusion of Hierarchical HotNet (permutation part) and NetCore (coreness normalization part). The algorithm highlighted the perturbed proteins which are close in the PPI network topology to proteins with high MS.ResultHighly perturbed proteins were more strongly associated with AD and other age‐related diseases (e.g.: AMD, Ischemic stroke) than proteins with high MS. Several pathways involved in AD were found to be associated to the perturbed proteins, such as complement and monocyte activation, or regulation of neurogenesis. Furthermore, the high overlap between proteins perturbed by GRF6019 and known AD‐associated proteins indicates the importance of indirect links between the GRF6019 treatment and AD.ConclusionNetwork propagation procedure highlights the importance of interactome‐based analysis in understanding complex plasma proteomics signals. These results suggest that the treatment of AD patients with a complex plasma fraction modulates AD‐related biological pathways.

Optimizing Graph Neural Network With Multiaspect Hilbert-Schmidt Independence Criterion.

The graph neural network (GNN) has demonstrated its superior power in various data mining tasks and has been widely applied in diversified fields. The core of GNN is the aggregation and combination functions, and mainstream GNN studies focus on the enhancement of these functions. However, GNNs face a common challenge, i.e., useless features contained in neighbor nodes may be integrated into the target node during the aggregation process. This leads to poor node embedding and undermines downstream tasks. To tackle this problem, this article proposes a novel GNN optimization framework GNN-MHSIC by introducing the nonparametric dependence method Hilbert-Schmidt independence criterion (HSIC) under the guidance of information bottleneck. HSIC is utilized to guide the information propagation among layers of a GNN from multiaspect views. GNN-MHSIC aims to achieve three main objectives: 1) minimizing the HSIC between the input features and the propagation layers; 2) maximizing the HSIC between the propagation layers and the ground truth; and 3) minimizing the HSIC between the propagation layers. With a multiaspect design, GNN-MHSIC can minimize the propagation of redundant information while preserving relevant information about the target node. We prove GNN-MHSIC's finite upper and lower bounds theoretically and evaluate it experimentally with four classic GNN models, including the graph convolutional network, the graph attention network (GAT), the heterogeneous GAT, and the heterogeneous graph (HG) propagation network on three widely used HGs. The results illustrate the usefulness and performance of GNN-MHSIC.

On the Group-Fairness-Aware Influence Maximization in Social Networks

The goal of influence maximization (IM) is to find a set of <inline-formula> <tex-math notation="LaTeX">$k$</tex-math> </inline-formula> nodes that maximize their influence spread over a social network. IM is a widely used model for information propagation in social networks. Although IM has been widely investigated in recent years, the fair propagation of information is still understudied. In this article, we consider the group-fairness-aware IM, where the objective is to ensure that information has been fairly distributed across different groups in the population. We extend the notion of group fairness in IM, introduced by Tsang <i>et al.</i>, by considering the speed of information propagation in different groups of a social network and formulating it as a new group-fairness metric. We propose a multiobjective metaheuristic (SetMOGWO), based on the multiobjective gray wolf optimizer, to improve the fair propagation of information in IM concerning various fairness metrics. Experimental results show that the proposed algorithm outperforms the previous work with respect to all introduced fairness metrics. We also carry out detailed experimental analyses on real-world networks and report interesting relationships between fairness concepts and network characteristics, including the price of ensuring fairness in social networks, the effect of social groups&#x2019; structure on fairness, and the dependence among fairness metrics. Especially, we show that traditional IM techniques often neglect smaller groups. On the other hand, we show that, with a low cost, it can be ensured that these groups receive a fair portion of resources in fair order.

Reactivity-Guided Depercolation Processes Determine Fracture Behavior in End-Linked Polymer Networks.

The fracture of polymer networks is tied to the molecular behavior of strands within the network, yet the specific molecular-level processes that determine the mechanical limits of a network remain elusive. Here, the question of reactivity-guided fracture is explored in otherwise indistinguishable end-linked networks by tuning the relative composition of strands with two different mechanochemical reactivities. Increasing the substitution of less mechanochemically reactive ("strong") strands into a network comprising more reactive ("weak") strands has a negligible impact on the fracture energy until the strong strand content reaches approximately 45%, at which point the fracture energy sharply increases with strong strand content. This aligns with the measured strong strand percolation threshold of 48 ± 3%, revealing that depercolation, or the loss of a percolated network structure, is a necessary criterion for crack propagation in a polymer network. Coarse-grained fracture simulations agree closely with the tearing energy trend observed experimentally, confirming that weak strand scissions dominate the failure until the strong strands approach percolation. The simulations further show that twice as many strands break in a mixture than in a pure network.

Misinformation Propagation in Online Social Networks: Game Theoretic and Reinforcement Learning Approaches

Misinformation in online social networks (OSNs) has been an ongoing problem, and it has been studied heavily over recent years. In this article, we use gamification to tackle misinformation propagation in OSNs. First, we construct a game based on the notion of cooperative games on graphs where the nodes of the social network are players. We use random regular networks and real networks in our simulations to show that the constructed game follows evolutionary dynamics and that the outcome of the game depends on the relation between the structural properties of the network and the benefit and cost variables defined in a cooperative game. Second, we create a game on the network level where the players control a set of nodes. We define agents whose goal is to maximize the total reward that we set up to be the number of nodes affected at the end of the game. We propose a deep reinforcement learning (RL) technique based on the multiagent deep deterministic policy gradient (MADDPG) algorithm. We test the proposed method along with well-known node selection algorithms and obtain promising results on different social networks.

Energy controls wave propagation in a neural network with spatial stimuli

Nervous system has distinct anisotropy and some intrinsic biophysical properties enable neurons present various firing modes in neural activities. In presence of realistic electromagnetic fields, non-uniform radiation activates these neurons with energy diversity. By using a feasible model, energy function is obtained to predict the growth of synaptic connections of these neurons. Distribution of average value of the Hamilton energy function vs. intensity of noisy disturbance can predict the occurrence of coherence resonance, which the neural activities show high regularity by applying noisy disturbance with moderate intensity. From physical viewpoint, the average energy value has similar role average power for the neuron. Non-uniform spatial disturbance is applied and energy is injected into the neural network, statistical synchronization factor is calculated to predict the network synchronization stability and wave propagation. The intensity for field coupling is adaptively controlled by energy diversity between adjacent neurons. Local energy balance will terminate further growth of the coupling intensity; otherwise, heterogeneity is formed in the network due to energy diversity. Furthermore, memristive channel current is introduced into the neuron model for perceiving the effect of electromagnetic induction and radiation, and a memristive neuron is obtained. The circuit implement of memristive circuit depends on the connection to a magnetic flux-controlled memristor into the mentioned neural circuit in an additive branch circuit. The connection and activation of this memristive neural network are controlled under external spatial electromagnetic radiation by capturing enough field energy. Continuous energy collection and exchange generate energy diversity and synaptic connection is created to regulate the synchronous firing patterns and energy balance.

Integrated systems biology approach identifies gene targets for endothelial dysfunction.

Endothelial dysfunction (ED) is critical in the development and progression of cardiovascular (CV) disorders, yet effective therapeutic targets for ED remain elusive due to limited understanding of its underlying molecular mechanisms. To address this gap, we employed a systems biology approach to identify potential targets for ED. Our study combined multi omics data integration, with siRNA screening, high content imaging and network analysis to prioritise key ED genes and identify a pro- and anti-ED network. We found 26 genes that, upon silencing, exacerbated the ED phenotypes tested, and network propagation identified a pro-ED network enriched in functions associated with inflammatory responses. Conversely, 31 genes ameliorated ED phenotypes, pointing to potential ED targets, and the respective anti-ED network was enriched in hypoxia, angiogenesis and cancer-related processes. An independent screen with 17 drugs found general agreement with the trends from our siRNA screen and further highlighted DUSP1, IL6 and CCL2 as potential candidates for targeting ED. Overall, our results demonstrate the potential of integrated system biology approaches in discovering disease-specific candidate drug targets for endothelial dysfunction.

Modeling Microscale Foam Propagation in a Heterogeneous Grain-Based Pore Network with the Pore-Filling Event Network Method

Foam flooding is an efficient and promising technology of enhanced oil recovery that significantly improves sweep efficiency of immiscible displacement processes by providing favorable mobility control on displacing fluids. Although the advantages in flexibility and efficiency are apparent, accurate prediction and effective control of foam flooding in field applications are still difficult to achieve due to the complexity in multiphase interactions. Also, conventional field-scale or mesoscale foam models are inadequate to simulate recent experimental findings in feasibility of foam injection in tight reservoirs. Microscale modeling of foam behavior has been applied to further connect those pore-scale interactions and mesoscale multiphase properties such as foam texture and the relative permeability of foam banks. Modification on a microscale foam model based on a pore-filling event network method is proposed to simulate its propagation in grain-based pore networks with varying degrees of heterogeneity. The impacts of foam injection strategy and oil-weakening phenomena are successfully incorporated. Corresponding microfluidic experiments are performed to validate the simulation results in dynamic displacement pattern as well as interfacial configuration. The proposed modeling method of foam propagation in grain-based networks successfully captures the effects of lamellae configurations corresponding to various foaming processes. The results of the simulation suggest that the wettability of rock has an impact on the relevance between reservoir heterogeneity and the formation of immobile foam banks, which supports the core idea of the recently proposed foam injection strategy in tight oil reservoirs with severe heterogeneity, that of focusing more on the IFT adjustment ability of foam, instead of arbitrarily pursuing high-quality strong foam restricted by permeability constraints.

Research on homogeneous information propagation in social network

In real network systems, there are numerous information propagation phenomena. However, the impact of homogeneity on information propagation is still not well understood. In this paper, we introduce a dynamic information propagation model named H-SIR. We also incorporate the propagation mechanism of content homogeneity and define the dynamic propagation rate based on users’ interest locking and fatigue level. Through simulation experiments, we demonstrate how homogeneous information spreads and evolves on the network. Furthermore, we compare the model with real data of COVID-19 information on the Sina-Weibo social platform and the traditional SIR model to jointly validate our model. Our results show that in social networks, if the themes of homogeneous information are diverse, even with an increasing amount of homogeneous information, it will hardly affect the propagation of this type of information. However, if users are recommended excessive homogeneous information, they will not focus entirely but instead display selective contact behavior. In the case of information fatigue, a type of information with a low degree of homogeneity will survive longer than a type of information with a high degree of homogeneity. Finally, we propose corresponding strategies to alleviate the phenomenon of information homogenization in social networks.

Linguistic Characteristics and Stylisation of Korean Cinema Based on the Social Network Analysis Method

Abstract Korean movies have consistently produced exceptional works that have triggered discussions and spread on social networks in recent years. In this paper, we take the movie language dissemination characteristics in social networks as an anchor point and combine the scale and structure of social networks to improve on the IC model to obtain a movie information dissemination model. This network model and the information cascade tandem are selected as the core of density peak clustering, and improvements are made for the updating rules as well as the propagation order of the multi-strategy labeling algorithm in the DCN algorithm. The characterization of movie language propagation in social networks is completed by using an improved algorithm with a multi-labeling idea. The overall stylistic evolution of Korean movies is examined, and then different language feature styles are compared according to social propagation. In the movie Memories of Murder, at 32 minutes and 23 seconds, which shows a rainy night on a muddy land and conveys the hardship of the protagonist’s situation through visual language, the propagation heat reaches 10.52. Similar to this sentence, the movie depicts a rainy night at 63 minutes and a crime scene at 108 minutes, with the propagation heat of the visual language reaching 19.48 and 23.48, respectively. Based on the propagation analysis of the social networks, the accurate capture of the unique impression of the movie language among the audience.

Integrating somatic mutation profiles with structural deep clustering network for metabolic stratification in pancreatic cancer: a comprehensive analysis of prognostic and genomic landscapes.

Pancreatic cancer is a globally recognized highly aggressive malignancy, posing a significant threat to human health and characterized by pronounced heterogeneity. In recent years, researchers have uncovered that the development and progression of cancer are often attributed to the accumulation of somatic mutations within cells. However, cancer somatic mutation data exhibit characteristics such as high dimensionality and sparsity, which pose new challenges in utilizing these data effectively. In this study, we propagated the discrete somatic mutation data of pancreatic cancer through a network propagation model based on protein-protein interaction networks. This resulted in smoothed somatic mutation profile data that incorporate protein network information. Based on this smoothed mutation profile data, we obtained the activity levels of different metabolic pathways in pancreatic cancer patients. Subsequently, using the activity levels of various metabolic pathways in cancer patients, we employed a deep clustering algorithm to establish biologically and clinically relevant metabolic subtypes of pancreatic cancer. Our study holds scientific significance in classifying pancreatic cancer based on somatic mutation data and may provide a crucial theoretical basis for the diagnosis and immunotherapy of pancreatic cancer patients.

A Novel RUL Prognosis Model Based on Counterpropagating Learning Approach

The aviation industry is one of the fastest-growing sectors and is crucial for both passenger transport and logistics. However, the high costs associated with maintenance, refurbishment, and overhaul (MRO) constitute one of the biggest challenges facing this industry. Motivated by the significant role that remaining useful life (RUL) prognostics can play in optimising MRO operations and saving lives, this paper proposes a novel data-driven RUL prognosis model based on counter propagation network principles. The proposed model introduces the recursive growing hierarchical self-organisation map (ReGHSOM) as a variant of SOM that can cluster multivariate time series with high correlations and hierarchical dependencies typically found in RUL datasets. Moreover, ReGHSOM is designed to allow this clustering to evolve dynamically at runtime without imposing constraints or prior assumptions on the hypothesis spaces of the architectures. The output of ReGHSOM is fed into the supervised learning layers of Grossberg to make the RUL prediction. The performance of the proposed model is comprehensively evaluated by measuring its learnability, evolution, and comparison with related work using standard statistical metrics. The results of this evaluation show that the model can achieve an average mean square error of 5.24 and an average score of 293 for the C-MPASS dataset, which are better results than most of the comparable works.

Comprehensive Simulation Framework for Space-Air-Ground Integrated Network Propagation Channel Research.

The space-air-ground integrated network (SAGIN) represents a pivotal component within the realm of next-generation mobile communication technologies, owing to its established reliability and adaptable coverage capabilities. Central to the advancement of SAGIN is propagation channel research due to its critical role in aiding network system design and resource deployment. Nevertheless, real-world propagation channel research faces challenges in data collection, deployment, and testing. Consequently, this paper designs a comprehensive simulation framework tailored to facilitate SAGIN propagation channel research. The framework integrates the open source QuaDRiGa platform and the self-developed satellite channel simulation platform to simulate communication channels across diverse scenarios, and also integrates data processing, intelligent identification, algorithm optimization modules in a modular way to process the simulated data. We also provide a case study of scenario identification, in which typical channel features are extracted based on channel impulse response (CIR) data, and recognition models based on different artificial intelligence algorithms are constructed and compared.

Multiscale variational autoencoder regressor for production prediction and energy saving of industrial processes

As modern industries increase in scale and integration, the industrial process data show more complex dynamic time variability. To solve the dynamic time-varying problem of industrial process data, a novel multiscale variational autoencoder (MSVAE) based Regressor (REG) (MSVAE-REG) is proposed for production prediction and energy saving. The encoder of the MSVAE is constructed by multiscale convolutional neural network (MSCNN) to extract the multiscale temporal dynamic features, including the overall and local trend features in the Gaussian latent space. Then, the decoder of the MSVAE recovers the input data to enhance the dynamic adaptability of multiscale features. Moreover, the REG utilizes the gated recurrent unit (GRU) to build a dynamic relationship between multiscale temporal dynamic features and the predicted output. Finally, the proposed model is verified on the propylene and ethylene production datasets. Compared with the back propagation (BP) network, the extreme learning machine (ELM), the radial basis function (RBF), the convolutional neural network (CNN) and the variational autoencoder (VAE), the MSVAE-REG achieves state-of-the-art results, with the prediction accuracy respectively reaching about 95% and 96%, providing a new judgment basis for optimizing the production process and saving energy.

Prediction of comprehensive dynamic performance for probability screen based on AR model-box dimension

In order to evaluate the comprehensive dynamic performance of probability screen and select the appropriate working conditions, a dynamic model of probability screen vibration system is established. Then, the calculation method of the dynamic characteristic parameters, based on the time series Auto Regression (AR) model of vibration test, is used. The relationship among the comprehensive dynamic characteristics, the screening efficiency and the box dimension of probability screen vibration system is analyzed, and Least Square Support Vector Machine (LSSVM), Generalized Regression Neural Network (GRNN) and Back Propagation Neural Network (BPNN) are used to predict the screening efficiency with box dimension. The analysis result shows that the screening efficiency, the stability, the response rapidity and the comprehensive dynamic characteristic of the system are all related to the box dimension of time series. As for the complexity of probability screen vibration system, it affects the comprehensive dynamic performance, and ultimately touches the screening efficiency of the probability screen; The best working conditions for the system are selected by the curve between box dimension and the working condition parameter; Taking box dimension as the only input variable, the prediction accuracy of the screening efficiency is high by using LSSVM,GRNN and BPNN methods, the prediction results are stable and reliable, and the box dimension can be used as a single input variable to predict the screening efficiency, it has the advantages of fewer input parameters, high prediction efficiency, and high prediction accuracy, which has great potential for expanding application space and further research value.

Transport causality knowledge-guided GCN for propagated delay prediction in airport delay propagation networks

Flight delays pose a worldwide challenge that significantly affect the safety and efficiency of air transportation systems. However, propagated delay prediction, as well as its causality among airport delay propagation networks, has not considered some crucial issues regarding spatiotemporal dependence and propagation relationships. Thus, this study proposes a transport causality knowledge-guided extended graph convolutional network (GCN) framework to tackle these issues. In particular, a causality knowledge-guided airport delay propagation network (ADPN) is developed using the second modified transfer entropy (SMTE) principle. Furthermore, a causality-embedded adjacency matrix is utilized by an extended GCN for propagated delay prediction. Comprehensive validations and results indicate that the proposed method benefits significantly from the causality knowledge, and increases the prediction performances up to 15.51%. Thus, transport causality is significant and efficient for understanding propagated delay features and airport delay propagation network characteristics.

Machine learning combining multi-omics data and network algorithms identifies adrenocortical carcinoma prognostic biomarkers.

Background: Rare endocrine cancers such as Adrenocortical Carcinoma (ACC) present a serious diagnostic and prognostication challenge. The knowledge about ACC pathogenesis is incomplete, and patients have limited therapeutic options. Identification of molecular drivers and effective biomarkers is required for timely diagnosis of the disease and stratify patients to offer the most beneficial treatments. In this study we demonstrate how machine learning methods integrating multi-omics data, in combination with system biology tools, can contribute to the identification of new prognostic biomarkers for ACC. Methods: ACC gene expression and DNA methylation datasets were downloaded from the Xena Browser (GDC TCGA Adrenocortical Carcinoma cohort). A highly correlated multi-omics signature discriminating groups of samples was identified with the data integration analysis for biomarker discovery using latent components (DIABLO) method. Additional regulators of the identified signature were discovered using Clarivate CBDD (Computational Biology for Drug Discovery) network propagation and hidden nodes algorithms on a curated network of molecular interactions (MetaBase™). The discriminative power of the multi-omics signature and their regulators was delineated by training a random forest classifier using 55 samples, by employing a 10-fold cross validation with five iterations. The prognostic value of the identified biomarkers was further assessed on an external ACC dataset obtained from GEO (GSE49280) using the Kaplan-Meier estimator method. An optimal prognostic signature was finally derived using the stepwise Akaike Information Criterion (AIC) that allowed categorization of samples into high and low-risk groups. Results: A multi-omics signature including genes, micro RNA's and methylation sites was generated. Systems biology tools identified additional genes regulating the features included in the multi-omics signature. RNA-seq, miRNA-seq and DNA methylation sets of features revealed a high power to classify patients from stages I-II and stages III-IV, outperforming previously identified prognostic biomarkers. Using an independent dataset, associations of the genes included in the signature with Overall Survival (OS) data demonstrated that patients with differential expression levels of 8 genes and 4 micro RNA's showed a statistically significant decrease in OS. We also found an independent prognostic signature for ACC with potential use in clinical practice, combining 9-gene/micro RNA features, that successfully predicted high-risk ACC cancer patients. Conclusion: Machine learning and integrative analysis of multi-omics data, in combination with Clarivate CBDD systems biology tools, identified a set of biomarkers with high prognostic value for ACC disease. Multi-omics data is a promising resource for the identification of drivers and new prognostic biomarkers in rare diseases that could be used in clinical practice.

Analysis of the factors influencing the promotion of stone sculpture art based on social media communication dynamics

Abstract This paper first focuses on the design elements and digital platforms of stone carving art and then discusses the dissemination of stone carving art on social media platforms. Secondly, a social media propagation dynamics modeling of stone sculpture art promotion is proposed, and the degree distribution of stone sculpture art promotion in the BA scale-free network is calculated by the social media network model and propagation evolution model. Finally, the kinetic analysis of social media on stone carving art promotion by different influencing factors and information dissemination stages is explored. The results show that when the user recognizes stone carving art as Ii (0) = 0.7 the larger the value Ii = (0), the greater the user’s propensity to spread stone carving art promotion information and the greater the rate of stone carving art promotion information dissemination. Through the in-depth combination of social media and stone carving art, this paper has important practice and significance for the protection and inheritance of stone carving art and for improving its influence and popularity.

GNR: A universal and efficient node ranking model for various tasks based on graph neural networks

Node ranking is an essential problem in complex networks, which aims to identify influential or central nodes in a graph. Existing methods for node ranking are either computationally expensive or suboptimal for different networks and multiple tasks. In this paper, we propose GNR, a graph neural network-based node ranking model that can sort nodes quickly and efficiently for any type of networks. To achieve this, GNR takes into account three factors to calculate the importance of nodes and train the model: degree, infection score, and dismantling score, which capture the local and global structure of the network and the effect of nodes on network propagation and dismantling. We conduct extensive experiments on 16 real networks, and compare GNR with other centrality-based node ranking models on three tasks: network dismantling, virus propagation, and information spreading. The results show that GNR performs better in most cases and validates the effectiveness and superiority of our model.

Variable neighbourhood search for connected graphs of fixed order and size with minimal spectral radius

In this study we consider connected graphs of fixed order n and size m that minimize the largest eigenvalue of the adjacency matrix, also known as the spectral radius. Such graphs are called minimizers. The motivation for this research lies in the fact that the spectral radius plays a significant role in modelling virus propagation in complex networks, in the sense that a smaller spectral radius ensures a better virus protection in the network modelled by the corresponding graph. We conjecture that vertex degrees of a minimizer are as equal as possible, i.e. belong to {⌊2m/n⌋,⌈2m/n⌉}. The conjecture is confirmed for graphs with at most 10 vertices by the total enumeration, and some classes of graphs are resolved theoretically. In general, exact determining of minimizers is quite difficult and computationally exhaustive, and thus we propose a long-scale variable neighbourhood search as an alternative approach. We employ this heuristic search for selected instances concerning graphs with at most 100 vertices, and as a result we always obtain a graph with required vertex degrees. The search efficiently results in solutions that are (in a precise sense) close to the optimal ones, i.e. to minimizers.