Multiple Types Of Interactions Research Articles

Collective relational inference for learning heterogeneous interactions

Interacting systems are ubiquitous in nature and engineering, ranging from particle dynamics in physics to functionally connected brain regions. Revealing interaction laws is of fundamental importance but also particularly challenging due to underlying configurational complexities. These challenges become exacerbated for heterogeneous systems that are prevalent in reality, where multiple interaction types coexist simultaneously and relational inference is required. Here, we propose a probabilistic method for relational inference, which possesses two distinctive characteristics compared to existing methods. First, it infers the interaction types of different edges collectively by explicitly encoding the correlation among incoming interactions with a joint distribution, and second, it allows handling systems with variable topological structure over time. We evaluate the proposed methodology across several benchmark datasets and demonstrate that it outperforms existing methods in accurately inferring interaction types. The developed methodology constitutes a key element for understanding interacting systems and may find application in graph structure learning.

‘If you make the people run away, you will starve’: the military significance of refugees during the Thirty Years’ War (1618–1648)

The Thirty Years’ War of 1618–1648 was the most devastating conflict in European history prior to the twentieth century and the greatest demographic crisis since the Black Death of the fourteenth century. Accordingly, many scholars have examined the ‘human cost’ of the Thirty Years’ War both in broader studies of the military and civilian experience of the conflict and with an increased focus on the multiple types of localised civil-military interactions which characterised the ‘small war’ or Kleinkrieg. This article takes these studies further, by examining how refugees – a neglected manifestation of the ‘human cost’ of the Thirty Years’ War – could in turn have a significant impact on how warfare was conducted locally and on the course of the conflict itself.

Design and synthesis of novel imidazole-chalcone derivatives as microtubule protein polymerization inhibitors to treat cervical cancer and reverse cisplatin resistance

Using the licochalcone moiety as a lead compound scaffold, 16 novel imidazole-chalcone derivatives were designed and synthesized as microtubule protein polymerization inhibitors. The proliferation inhibitory activities of the derivatives against SiHa (human cervical squamous cell carcinoma), C-33A (human cervical cancer), HeLa (human cervical cancer), HeLa/DDP (cisplatin-resistant human cervical cancer), and H8 (human cervical epithelial immortalized) cells were evaluated. Compound 5a exhibited significant anticancer activity with IC50 values ranging from 2.28 to 7.77 μM and a resistance index (RI) of 1.63, while showing minimal toxicity to normal H8 cells. When compound 5a was coadministered with cisplatin, the RI of cisplatin to HeLa/DDP cells decreased from 6.04 to 2.01, while compound 5a enhanced the fluorescence intensity of rhodamine 123 in HeLa/DDP cells. Further studies demonstrated that compound 5a arrested cells at the G2/M phase, induced apoptosis, reduced colony formation, inhibited cell migration, and inhibited cell invasion. Preliminary mechanistic studies revealed that compound 5a decreased the immunofluorescence intensity of α-/β-tubulin in cancer cells, reduced the expression of polymerized α-/β-tubulin, and increased the expression of depolymerized α-/β-tubulin. Additionally, the molecular docking results demonstrate that compound 5a can interact with the tubulin colchicine binding site and generate multiple types of interactions. These results suggested that compound 5a has anticancer effects and significantly reverses cervical cancer resistance to cisplatin, which may be related to its inhibition of microtubule and P-glycoprotein (P-gp) activity.

Quintuple-based Representation Learning for Bipartite Heterogeneous Networks

Recent years have seen rapid progress in network representation learning, which removes the need for burdensome feature engineering and facilitates downstream network-based tasks. In reality, networks often exhibit heterogeneity, which means there may exist multiple types of nodes and interactions. Heterogeneous networks raise new challenges to representation learning, as the awareness of node and edge types is required. In this paper, we study a basic building block of general heterogeneous networks, the heterogeneous networks with two types of nodes. Many problems can be solved by decomposing general heterogeneous networks into multiple bipartite ones. Recently, to overcome the demerits of non-metric measures used in the embedding space, metric learning-based approaches have been leveraged to tackle heterogeneous network representation learning. These approaches first generate triplets of samples, in which an anchor node, a positive counterpart and a negative one co-exist, and then try to pull closer positive samples and push away negative ones. However, when dealing with heterogeneous networks, even the simplest two-typed ones, triplets cannot simultaneously involve both positive and negative samples from different parts of networks. To address this incompatibility of triplet-based metric learning, in this paper, we propose a novel quintuple-based method for learning node representations in bipartite heterogeneous networks. Specifically, we generate quintuples that contain positive and negative samples from two different parts of networks. And we formulate two learning objectives that accommodate quintuple-based learning samples, a proximity-based loss that models the relations in quintuples by sigmoid probabilities, and an angular loss that more robustly maintains similarity structures. In addition, we also parameterize feature learning by using one-dimensional convolution operators around nodes’ neighborhoods. Compared with eight methods, extensive experiments on two downstream tasks manifest the effectiveness of our approach.

Mechanisms of Pore-Clogging Using a High-Resolution CFD-DEM Colloid Transport Model

Colloidal transport and clogging in porous media is a phenomenon of critical importance in many branches of applied sciences and engineering. It involves multiple types of interactions that span from the sub-colloid scale (electrochemical interactions) up to the pore-scale (bridging), thus challenging the development of representative modelling. So far published simulation results of colloidal or particulate transport are based on either reduced set of forces or spatial dimensions. Here we present an approach enabling to overcome both computational and physical limitations posed by a problem of 3D colloidal transport in porous media. An adaptive octree mesh is introduced to a coupled CFD and DEM method while enabling tracking of individual colloids. Flow fields are calculated at a coarser scale throughout the domain, and at fine-scale around colloids. The approach accounts for all major interactions in such a system: elastic, electrostatic, and hydrodynamic forces acting between colloids, as well as colloids and the collector surface. The method is demonstrated for a single throat model made of four spherical segments, and the impact of clogging is reported in terms of the evolution of the critical path diameter for percolation and permeability. We identified four stages of clogging development depending on position and time of individual colloid entrapment, which in turn correlates to a cluster evolution and local transport.

Prediction of multiple types of drug interactions based on multi-scale fusion and dual-view fusion.

Potential drug-drug interactions (DDI) can lead to adverse drug reactions (ADR), and DDI prediction can help pharmacy researchers detect harmful DDI early. However, existing DDI prediction methods fall short in fully capturing drug information. They typically employ a single-view input, focusing solely on drug features or drug networks. Moreover, they rely exclusively on the final model layer for predictions, overlooking the nuanced information present across various network layers. To address these limitations, we propose a multi-scale dual-view fusion (MSDF) method for DDI prediction. More specifically, MSDF first constructs two views, topological and feature views of drugs, as model inputs. Then a graph convolutional neural network is used to extract the feature representations from each view. On top of that, a multi-scale fusion module integrates information across different graph convolutional layers to create comprehensive drug embeddings. The embeddings from the two views are summed as the final representation for classification. Experiments on two real-world datasets demonstrate that MSDF achieves higher accuracy than state-of-the-art methods, as the dual-view, multi-scale approach better captures drug characteristics.

Dual species interaction and ecological community stability.

How diverse species coexist in nature remains a challenging issue that is not yet resolved in ecology. The traditional approach to tackling this problem uses an ecological community network comprising various biological interaction links between species, such as predator-prey, mutualism, and competition. However, in nature, the interaction between any species pair is not limited to a singular interaction; instead, various interactions occur mostly in two ways, such as competition/facilitation in plants, mutualism/antagonism in consumer-resource mutualisms, and reciprocal predation. Here, using an ecological community model, I show that such so-called dual interactions play a key role in stabilizing ecological communities. Theory predicts that dual interactions can stabilize ecological communities through the balance of positive and negative effects, which behave as if the interactions disappear. Communities with dual interactions are inherently more stable than a classical random community with multiple types of singular interactions, suggesting that dual interactions are more widespread than expected in nature and help to maintain ecological communities.

Adsorption of perfluoroalkyl substances on deep eutectic solvent-based amorphous metal-organic framework: Structure and mechanism

Perfluoroalkyl substances (PFASs) are a class of emerging organic pollutants characterized by high toxicity, environmental persistence, and widespread detection in water sources. The removal of PFASs from water is a matter of global concern, given their detrimental impact on both the environment and public health. Many commonly used PFAS adsorbents demonstrate limited adsorption capacities and/or slow adsorption kinetics. Therefore, there is an urgent need for the development of efficient adsorbents. For the first time, this work systematically investigated the performance of a deep eutectic solvent (DES)-based amorphous metal-organic framework (MOF) for the adsorption of PFASs with different carbon-chain lengths under the state of the mixture in aquatic environments. The adsorption mechanism was probed by a suite of adsorption kinetics studies, adsorption isotherm profiling, spectral characterization, and ab initio molecular dynamics (AIMD) simulations, revealing that PFAS adsorption is driven by synergistic capturing effects including acid/base coordination, CF-π (carbon-fluorine-π), hydrogen bonding, and hydrophobic interactions. Furthermore, the adsorption processes of short-chain and long-chain targets were found to involve different rate-controlling steps and interaction sites. Hydrophobic interactions facilitated the swift arrival of long-chain PFASs at the coordinatively interacting sites between carboxyl termini and Lewis acid Zr unsaturated sites, thanks to their lower reaction barriers. On the other hand, the adsorption of short-chain PFASs primarily relied on a Zr hydroxyl-based ligand exchange force, which would take place at Brønsted acid sites. The existence of massive structural disorder in amorphous UiO-66 led to the development of larger pores, thus improving the accessibility of abundant adsorption sites and facilitating adsorption and diffusion. The presence of multiple types of interactions and flexible structure in defect-rich amorphous UiO-66 significantly increased the exposure of functional groups to the adsorbates. Additionally, this material possessed outstanding regeneration efficiency and outperformed other MOF-based adsorbents with high affinity for targets. It enhances our understanding of the adsorption performances and mechanisms of amorphous materials toward PFASs, thereby paving the way for designing more efficient PFAS adsorbents.

The structure and robustness of ecological networks with two interaction types.

Until recently, most ecological network analyses investigating the effects of species' declines and extinctions have focused on a single type of interaction (e.g. feeding). In nature, however, diverse interactions co-occur, each of them forming a layer of a 'multilayer' network. Data including information on multiple interaction types has recently started to emerge, giving us the opportunity to have a first glance at possible commonalities in the structure of these networks. We studied the structural features of 44 tripartite ecological networks from the literature, each composed of two layers of interactions (e.g. herbivory and pollination), and investigated their robustness to species losses. Considering two interactions simultaneously, we found that the robustness of the whole community is a combination of the robustness of the two ecological networks composing it. The way in which the layers of interactions are connected to each other affects the interdependence of their robustness. In many networks, this interdependence is low, suggesting that restoration efforts would not automatically propagate through the whole community. Our results highlight the importance of considering multiple interactions simultaneously to better gauge the robustness of ecological communities to species loss and to more reliably identify key species that are important for the persistence of ecological communities.

Assessing mutualistic metacommunity capacity by integrating spatial and interaction networks

We develop a spatially realistic model of mutualistic metacommunities that exploits the joint structure of spatial and interaction networks. Assuming that all species have the same colonisation and extinction parameters, this model exhibits a sharp transition between stable non-null equilibrium states and a global extinction state. This behaviour allows defining a threshold on colonisation/extinction parameters for the long-term metacommunity persistence. This threshold, the ‘metacommunity capacity’, extends the metapopulation capacity concept and can be calculated from the spatial and interaction networks without needing to simulate the whole dynamics. In several applications we illustrate how the joint structure of the spatial and the interaction networks affects metacommunity capacity. It results that a weakly modular spatial network and a power-law degree distribution of the interaction network provide the most favourable configuration for the long-term persistence of a mutualistic metacommunity. Our model that encodes several explicit ecological assumptions should pave the way for a larger exploration of spatially realistic metacommunity models involving multiple interaction types.

A Malicious Information Traceability Model Based on Neighborhood Similarity and Multiple Types of Interaction

A Malicious Information Traceability Model Based on Neighborhood Similarity and Multiple Types of Interaction

Drug Encapsulation via Peptide-Based Polyelectrolyte Complexes.

Peptide-based polyelectrolyte complexes are biocompatible materials that can encapsulate molecules with different polarities due to their ability to be precisely designed. Here we use UV-Vis spectroscopy, fluorescence microscopy, and infrared spectroscopy to investigate the encapsulation of model drugs, doxorubicin (DOX) and methylene blue (MB) using a series of rationally designed polypeptides. For both drugs, we find an overall higher encapsulation efficiency with sequences that have higher charge density, highlighting the importance of ionic interactions between the small molecules and the peptides. However, comparing molecules with the same charge density, illustrated that the most hydrophobic sequence pairs had the highest encapsulation of both DOX and MB molecules. The phase behavior and stability of DOX-containing complexes did not change compared to the complexes without drugs. However, MB encapsulation caused changes in the stabilities of the complexes. The sequence pair with the highest charge density and hydrophobicity had the most dramatic increase in stability, which coincided with a phase change from liquid to solid. This study illustrates how multiple types of molecular interactions are required for efficient encapsulation of poorly soluble drugs and provides insights into the molecular design of delivery carriers.

Multi-dimensional opinion chaos synchronization based on multi-layer complex networks

Abstract In complex networks, the interactions of opinions are significantly affected by network structure. Moreover, multi-layer networks, which contain multiple subsystems and different types of interactions, are considered to have a stronger ability to simulate various real-world systems than traditional single-layer complex networks. Based on this motivation, this article attempts to show how to achieve complete synchronization of multi-dimensional opinions in a multi-layer network topology where the opinions of nodes have Rössler chaotic dynamics. A first step in investigating the synchronization of multi-layer complex networks is to propose a multi-dimensional opinion evolution model. In this article, the multi-layer complex network model combining the Rössler oscillation system is built, which provides us a tool for insight into the nodes complexity in multi-layer networks. Furthermore, a general criterion for globally exponentially asymptotic synchronization of multi-layer complex networks is provided using multiple Lyapunov stability theories. We illustrate the reliability of the model and the correctness of the criterion by Runge-Kutta (RK4) simulation. These findings shed new light on some interesting phenomena in society and help decision-makers predict the prevalence of hot topics and public opinions.

Practical guidelines and the EMLN R package for handling ecological multilayer networks

Abstract Network analysis provides a powerful framework to study the complexity underlying the structure, dynamics and function of ecological systems. By now, analysing single‐layered networks is a common practice with clear guidelines and well‐established computational tools. Nevertheless, ecological communities are multilayered because they vary over space and time and contain multiple types of interactions. In recent years, the analysis of ecological multilayer networks (EMLNs) has allowed researchers to include such multilayered complexity. However, there is a paucity of practical guidelines and standardised tools to handle EMLN data, even before downstream analysis. In this article, we accomplish three objectives: (1) We provide practical guidelines for handling EMLN data. (2) We developed the EMLN R package to standardise the workflow of creating, storing and working with EMLN objects. The package also enables conversion between multiple formats for downstream analysis with other standard packages. (3) We provide EMLN data sets for research and training purposes. A dedicated website with explanations, detailed examples and code accompanies our paper. This website is a gateway for novice and experienced network ecologists who want to include EMLNs in their research. By simplifying the analysis of multilayer networks and promoting standardised approaches, we facilitate the analysis of EMLNs. This paves the road towards gaining deeper insights into functioning and resilience of natural ecosystems.

The global risk of infectious disease emergence from giant land snail invasion and pet trade

BackgroundPathogen outbreaks mostly originate from animals, but some species are more likely to trigger epidemics. The giant land snail (Lissachatina fulica) is a widespread invader, a popular exotic pet, and a notorious vector of the rat lungworm, causing eosinophilic meningitis in humans. However, a comprehensive assessment of the risks of disease outbreak associated with this species is lacking.MethodsWe assessed and mapped the risk of disease transmission associated with the invasion and pet trade of L. fulica. First, we conducted a review of the scientific literature to list all known L. fulica parasites and pathogens and query host–pathogen databases to identify their potential mammalian hosts. Then, to assess the potential for L. fulica to spread globally, we modelled its suitable climatic conditions and tested whether, within climatically suitable areas, the species tended to occur near humans or not. Finally, we used social media data to map L. fulica possession as an exotic pet and to identify human behaviours associated with increased risk of disease transmission.ResultsLissachatina fulica can carry at least 36 pathogen species, including two-thirds that can infect humans. The global invasion of L. fulica is climatically limited to tropical areas, but the species is strongly associated with densely populated areas where snails are more likely to enter in contact with humans. In temperate countries, however, climatic conditions should prevent L. fulica's spread. However, we show that in Europe, giant snails are popular exotic pets and are often handled with direct skin contact, likely increasing the risk of pathogen transmission to their owners.ConclusionsIt is urgent to raise public awareness of the health risks associated with L. fulica in both tropical countries and Europe and to regulate its trade and ownership internationally. Our results highlight the importance of accounting for multiple types of human-wildlife interactions when assessing risks of infectious disease emergence. Furthermore, by targeting the species most likely to spread pathogens, we show that it is possible to rapidly identify emerging disease risks on a global scale, thus guiding timely and appropriate responses.Graphical

An Exploratory Application of Multilayer Networks and Pathway Analysis in Pharmacogenomics.

Over the years, network analysis has become a promising strategy for analysing complex system, i.e., systems composed of a large number of interacting elements. In particular, multilayer networks have emerged as a powerful framework for modelling and analysing complex systems with multiple types of interactions. Network analysis can be applied to pharmacogenomics to gain insights into the interactions between genes, drugs, and diseases. By integrating network analysis techniques with pharmacogenomic data, the goal consists of uncovering complex relationships and identifying key genes to use in pathway enrichment analysis to figure out biological pathways involved in drug response and adverse reactions. In this study, we modelled omics, disease, and drug data together through multilayer network representation. Then, we mined the multilayer network with a community detection algorithm to obtain the top communities. After that, we used the identified list of genes from the communities to perform pathway enrichment analysis (PEA) to figure out the biological function affected by the selected genes. The results show that the genes forming the top community have multiple roles through different pathways.

Modeling and monitoring multilayer attributed weighted directed networks via a generative model

As data with network structures are widely seen in diverse applications, the modeling and monitoring of network data have drawn considerable attention in recent years. When individuals in a network have multiple types of interactions, a multilayer network model should be considered to better characterize its behavior. Most existing network models have concentrated on characterizing the topological structure among individuals, and important attributes of individuals are largely disregarded in existing works. In this article, first, we propose a unified static Network Generative Model (static-NGM), which incorporates individual attributes in network topology modeling. The proposed model can be utilized for a general multilayer network with weighted and directed edges. A variational expectation maximization algorithm is developed to estimate model parameters. Second, to characterize the time-dependent property of a network sequence and perform network monitoring, we extend the static-NGM model to a sequential version, namely, the sequential-NGM model, with the Markov assumption. Last, a sequential-NGM chart is developed to detect shifts and identify root causes of shifts in a network sequence. Extensive simulation experiments show that considering attributes improves the parameter estimation accuracy and that the proposed monitoring method also outperforms the three competitive approaches, static-NGM chart, score test-based chart (ST chart) and Bayes factor-based chart (BF chart), in both shift detection and root cause diagnosis. We also perform a case study with Enron E-mail data; the results further validate the proposed method.

Understanding Learning Engagement with User-Centered Human-Computer Interaction in a Multimodal Online Learning Environment

Multimodal online learning environment improves learning experience through different modalities such as visual, auditory, and kinesthetic interactions. Multimodal learning analytics (MMLA) with multiple biosensors provides a way to overcome the challenge of analyzing the multiple interaction types simultaneously. Galvanic skin response/electrodermal activity (GSR/EDA), eye tracking and facial expression were used to measure the learning interaction in a multimodal online learning environment. iMotions and R software were used to post-process and analyze the time-synchronized biosensor data. GSR/EDA, eye tracking and facial expression showed real-time cognitive, emotional, and visual learning engagement for each interaction type. There is a tremendous potential for using MMLA with multiple biosensors to understand learning engagement in a multimodal online learning environment was shown in this study.

Cascading Residual Graph Convolutional Network for Multi-Behavior Recommendation

Multi-behavior recommendation exploits multiple types of user-item interactions, such as view and cart , to learn user preferences and has demonstrated to be an effective solution to alleviate the data sparsity problem faced by the traditional models that often utilize only one type of interaction for recommendation. In real scenarios, users often take a sequence of actions to interact with an item, in order to get more information about the item and thus accurately evaluate whether an item fits their personal preferences. Those interaction behaviors often obey a certain order, and more importantly, different behaviors reveal different information or aspects of user preferences towards the target item. Most existing multi-behavior recommendation methods take the strategy to first extract information from different behaviors separately and then fuse them for final prediction. However, they have not exploited the connections between different behaviors to learn user preferences. Besides, they often introduce complex model structures and more parameters to model multiple behaviors, largely increasing the space and time complexity. In this work, we propose a lightweight multi-behavior recommendation model named Cascading Residual Graph Convolutional Network ( CRGCN for short) for multi-behavior recommendation, which can explicitly exploit the connections between different behaviors into the embedding learning process without introducing any additional parameters (with comparison to the single-behavior based recommendation model). In particular, we design a cascading residual graph convolutional network (GCN) structure, which enables our model to learn user preferences by continuously refining the embeddings across different types of behaviors. The multi-task learning method is adopted to jointly optimize our model based on different behaviors. Extensive experimental results on three real-world benchmark datasets show that CRGCN can substantially outperform the state-of-the-art methods, achieving 24.76%, 27.28%, and 25.10% relative gains on average in terms of HR@K (K = {10,20,50,80}) over the best baseline across the three datasets. Further studies also analyze the effects of leveraging multi-behaviors in different numbers and orders on the final performance.

A positive complexity-stability relationship emerges in pollinator-plant-consumer tripartite networks disturbed by plant invasion

The relationship between network structure and system stability has been debated for decades. Recent studies have found that species invasions are able to change the structure-stability relationship in mutualistic or antagonistic bipartite networks. However, whether and how species invasion can alter the structure-stability in tripartite networks with both mutualistic and antagonistic interactions remain largely unexplored. Here, we establish a patch-occupancy dynamic (including species colonization-extinction processes) model to explore the effect of plant invasion (e.g., fragrant eupatorium herb invades native plant community in southern Yunnan, China) on the relationship between network structure (i.e., connectance, network size, modularity and nestedness) and community stability (robustness to invasion) in a pollinator-plant-consumer tripartite network (e.g., Lepidoptera insect-herbaceous vegetation-phytophagous insect tripartite network). Theoretical results show that nestedness and modularity have opposite effects on community stability. Specifically, our work reveals that, in both mutualistic and antagonistic bipartite networks, the effect of modularity on community stability is positive, whereas the effect of nestedness is negative. Moreover, increasing network size increases system stability. Finally, we find a positive connectance-stability relationship in both mutualistic and antagonistic bipartite networks. Overall, our patch-dynamic framework provides new insights into how species invasion affects the stability of tripartite networks with multiple interaction types.