Rewiring of molecular interactions under different conditions causes different phenotypic responses. Differential Network Analysis (also indicated as DNA) aims to investigate the rewiring of gene and protein networks. DNA algorithms combine statistical learning and graph theory to explore the changes in the interaction patterns starting from experimental observation. Despite there exist many methods to model rewiring in networks, we propose to use age and gender factors to guide rewiring algorithms. We present a novel differential network analysis method that consider the differential expression of genes by means of sex and gender attributes. We hypothesise that the expression of genes may be represented by using a non-gaussian process. We quantify changes in nonparametric correlations between gene pairs and changes in expression levels for individual genes. We apply our method to identify the differential networks between males and females in public expression datasets related to mellitus diabetes in liver tissue. Results show that this method can find biologically relevant differential networks.
Networks provide an understandable and, in the case of small size, visualizable representation of data, which allows us to obtain essential information about the relationships between pairs of nodes, e.g., their distances. In visualization, networks have an alternative two-dimensional vector representation to which various machine-learning methods can be applied. More generally, networks can be transformed into a low-dimensional space using so-called embedding methods, which bridge the gap between network analysis and traditional machine learning by creating numerical representations that capture the essence of the network structure. In this article, we present a new embedding method that uses non-symmetric dependency to find the distance between nodes and applies an iterative procedure to find a satisfactory distribution of nodes in space. For dimension 2 and the visualization of the result, we demonstrate the method’s effectiveness on small networks. For higher dimensions and several larger networks, we present the results of two experiments comparing our results with two well-established methods in the research community, namely node2vec and DeepWalk. The first experiment focuses on a qualitative comparison of the methods, while the second focuses on applying and comparing the classification results to embeddings in a higher dimension. Although the presented method does not outperform the two chosen methods, its results are still comparable. Therefore, we also explain the limitations of our method and a possible way to overcome them.
People who inject drugs (PWID) are often members of HIV/AIDS risk networks, where individuals engage in sexual and injection risk behavior. Engagement in HIV care is important for this population. Low socioeconomic status, stigmatization, and lack of access to medical care often complicate successful engagement in the HIV cascade of care for PWID. This study investigates how individual's attitudes about how much control they have over HIV/AIDS risk in their life (i.e., locus of control and self-blame) affect health-seeking behavior in PWID participants and their community members. We applied causal inference methodology to PWID HIV risk networks ascertained from the Social Factors and HIV Risk Study (SFHR) conducted between 1991 and 1993 in Bushwick, Brooklyn, New York. We estimated protective disseminated effects of attitudes toward HIV/AIDS on the health-seeking behaviors of others in the PWID community. In other words, a positive attitude toward controlling HIV/AIDS can improve the health-seeking behavior of other members of the community who report a pessimistic attitude toward HIV/AIDS control. Given this finding, we also discuss potential network interventions to improve health-seeking behavior among both PWID individuals who receive the intervention and others in the PWID network informed by our analysis of disseminated effects.
Topological Data Analysis is a field of great interest in many applications such as finance or neuroscience. The goal of the present paper is to propose a novel approach to building simplicial complexes that capture the multiway ordered interactions in the components of high-dimensional time series using the theory of Signatures. Signatures represent one of the most powerful transforms for extracting group-wise structural features and we put them to work in the task of discovering statistically meaningful simplices from a complex that we estimate sequentially. Numerical experiments on an fMRI dataset illustrates the efficiency and relevance of our approach.
Uniform sampling from the set $\mathcal{G}(\mathbf{d})$ of graphs with a given degree-sequence $\mathbf{d} = (d_1, \dots, d_n) \in \mathbb N^n$ is a classical problem in the study of random graphs. We consider an analogue for temporal graphs in which the edges are labeled with integer timestamps. The input to this generation problem is a tuple $\mathbf{D} = (\mathbf{d}, T) \in \mathbb N^n \times \mathbb N_{>0}$ and the task is to output a uniform random sample from the set $\mathcal{G}(\mathbf{D})$ of temporal graphs with degree-sequence $\mathbf{d}$ and timestamps in the interval $[1, T]$. By allowing repeated edges with distinct timestamps, $\mathcal{G}(\mathbf{D})$ can be non-empty even if $\mathcal{G}(\mathbf{d})$ is, and as a consequence, existing algorithms are difficult to apply. We describe an algorithm for this generation problem which runs in expected time $O(M)$ if $\Delta^{2+\epsilon} = O(M)$ for some constant $\epsilon > 0$ and $T - \Delta = \Omega(T)$ where $M = \sum_i d_i$ and $\Delta = \max_i d_i$. Our algorithm applies the switching method of McKay and Wormald $[1]$ to temporal graphs: we first generate a random temporal multigraph and then remove self-loops and duplicated edges with switching operations which rewire the edges in a degree-preserving manner.
Given the rising prevalence of disinformation and fake news online, the detection of fake news in social media posts has become an essential task in the field of social network analysis and NLP. In this paper, we propose a fake detection model named, FakEDAMR that encodes textual content using the Abstract Meaning Representation (AMR) graph, a semantic representation of natural language that captures the underlying meaning of a sentence. The graphical representation of textual content holds longer relation dependency in very few distances. A new fake news dataset, FauxNSA, has been created using tweets from the Twitter platform related to ‘Nupur Sharma’ and ‘Agniveer’ political controversy. We embed each sentence of the tweet using an AMR graph and then use this in combination with textual features to classify fake news. Experimental results on publicly and proposed datasets with two different sets show that adding AMR graph features improves F1-score and accuracy significantly. (Code and Dataset: https://github.com/shubhamgpt007/FakedAMR )
Graph Neural Network (GNN) research has produced strategies to modify a graph's edges using gradients from a trained GNN, with the goal of network design. However, the factors which govern gradient-based editing are understudied, obscuring why edges are chosen and if edits are grounded in an edge's importance. Thus, we begin by analyzing the gradient computation in previous works, elucidating the factors that influence edits and highlighting the potential over-reliance on structural properties. Specifically, we find that edges can achieve high gradients due to structural biases, rather than importance, leading to erroneous edits when the factors are unrelated to the design task. To improve editing, we propose ORE, an iterative editing method that (a) edits the highest scoring edges and (b) re-embeds the edited graph to refresh gradients, leading to less biased edge choices. We empirically study ORE through a set of proposed design tasks, each with an external validation method, demonstrating that ORE improves upon previous methods by up to 50
Network visualization is an important tool for extracting information from the structure and configuration of a network, especially when the network includes weighted edges and nodes with attribute information. Previous studies have demonstrated an effective visualization method that projects the network onto a spherical surface. In this work, we extend this method, known as Uniform Placement of Alters on Spherical Surface (U-PASS), to general weighted networks. This extension enables the uniform distribution of network nodes across multiple layers of concentric spheres. In addition, we discuss the lower bound of a criterion called generalized spherical cap discrepancy, which is used to evaluate the uniformity of node distribution on a collapsed spherical surface.
This study delves into the privacy risks associated with user interactions in complex networks such as those generated on social media platforms. In such networks, potentially sensitive information can be extracted and/or inferred from explicitly user-generated content and its (often uncontrolled) dissemination. Hence, this preliminary work first studies an unsupervised model generating a privacy risk score for a given user, which considers both sensitive information released directly by the user and content propagation in the complex network. In addition, a supervised model is studied, which identifies and incorporates features related to privacy risk. The results of both multi-class and binary privacy risk classification for both models are presented, using the Twitter platform as a scenario, and a publicly accessible purpose-built dataset.
The disruption of transportation systems caused by natural hazards in one region can have significant consequences on the distribution of agricultural products and their export. In various regions of the world, climate change is expected to increase the likelihood of multiple natural hazards, such as landslides or floods. Being able to model how perturbations to transportation networks affect critical export routes is an important step toward making the system more resilient. Here, we analyze how disruptions to the Brazilian soybeans transportation network would impact export economics. We show that the impact to the Brazilian market can be important, with most of the main routes showing an impact of more than 10% on costs. This in turn can have a significant impact on the worldwide markets. We also show that mitigation measures can and should be taken to adapt to the network weaknesses, especially in the face of climate change.
This paper presents an adaptive network model in the context of joint action and social bonding. Exploration of mechanisms for mental and social network models are presented, specifically focusing on adaptation by bonding based on homophily and Hebbian learning during joint rhythmic action. The paper provides a comprehensive explanation of these concepts and their role in controlled adaptation within illustrative scenarios.
The Electronic messaging is a widely used network application, and the user’s authentication is a necessary aspect. Commonly employed approaches include PKI and S/MIME encryption protocols for email, but have many security threats, such as EFAIL and attack MITM attack. The innovative property of blockchain minimizes these threats and provides decentralized sensitive operations along with high security. It terminates the need of trusted intermediaries. Smart contracts are implementation of second generation of blockchain technology. The objective of our work is to provide a secure peer to peer messaging solution. In this paper, we explain how blockchain would ensure secure communications, and also, we have designed a model for messaging system that ensures good performance and data security which verifies the user identities and their public keys, and also validate the user certificate. The proposed messaging system is decentralized in nature and allows secure messages exchange.
We study the main decomposition approaches (primal, dual and primal–dual) for a distributed optimization problem from a dynamical system perspective where the couplings among the variables of the optimization problem are described by an undirected, unweighted hypergraph. We conduct stability analysis for the respective dynamical systems of the decomposition cases by using non linear decentralized control theoretic techniques and spectral properties of the respective communication matrices, i.e., the incidence and the Laplacian matrices of the hypergraph. Finally, we provide numerical simulations under a specific coalitional setting that demonstrate the superiority of the hypergraph compared to its respective graph analogue, the clique expansion graph, for the given decomposition algorithms in terms of convergence rate and information transmission efficiency.
Integral to the problem of detecting communities through graph clustering is the expectation that they are “well-connected”. Surprisingly, we find that the output of multiple clustering approaches–the Leiden algorithm with either the Constant Potts Model or modularity as quality function, Iterative K-Core Clustering, Infomap, and Markov Clustering–include communities that fail even a mild requirement for well-connectedness. As a remediation strategy, we have developed the “Connectivity Modifier” (CM), which iteratively removes small edge cuts and re-clusters until all communities detected are well-connected. Results from real-world networks with up to 75,025,194 nodes illustrate how CM enables additional insights into community structure within networks, while results on synthetic networks show that the CM algorithm improves accuracy in recovering true communities. Our study also raises questions about the “clusterability” of networks and mathematical models of community structure.
A network autocorrelation model may be embedded within a hierarchical data structure involving a complex network, when the peer effect (also referred to as social influence or contagion) is believed to act between units (e.g., hospitals) above the level at which data is observed. We develop two hierarchical network autocorrelation models to represent peer effects between hospitals when modeling individual outcomes of the patients who attend those hospitals. We use a Bayesian approach for model estimation and assess the performance of the models and sensitivity of results to different prior distributions using a simulation study. We construct a United States New England region patient-sharing hospital network and apply our models and Bayesian estimation procedures to study the diffusion of the adoption of robotic surgery and hospital-peer effects in the quality of patients outcomes using a cohort of United States Medicare beneficiaries in 2016 and 2017.
We present two new seeding strategies for the Influence Maximization Problem for Viral Marketing, based on graph connectivity and spectral graph theory. Specifically, the first approach CVSP uses the cut vertices and the separation pairs as the starting seeds. The second approach ER uses the vertex ranking based on the effective resistance values of the incident edges. CVSP and ER are efficient, and can be implemented in linear and near linear time, respectively. Experiments using the Independent Cascade diffusion model with real-world data sets show that our new seeding strategies perform significantly better than the existing methods, such as centrality measures, k-core and the state-of-the-art IMM, in particular for the scale-free networks with globally sparse, locally dense clusters with small diameters, in the final influence spread. Moreover, visual analysis enables more refined comparison between the methods, demonstrating that our methods have more globally wide influence spread pattern than other methods with locally dense influence spread pattern.
Social media and online data pose challenges in information mining, network analysis, opinion mining, and combating misinformation. However, no previous work has been able to apply knowledge graph (KG) and contextual focal structure analysis (CFSA) on multisource data to study situational awareness in public discussion and establish information propagation such as the Belt and Road Initiative (BRI). This research uses multisource data, a knowledge graph model, and a CFSA, which we term KG-CFSA. We extract entities and topics from documents and correlate them with third-party data sources such as Wikidata and Diffbot. We establish relationships using a Cartesian product merge function to develop a graph model. The merge function uses search algorithms and pairwise matching to establish relationships. The model is divided into three instances: document-entity, document-document, and topic-topic. For the document-document instance, we used topics and entities and topic overlaps to establish a relationship while we used co-occurrence for the topic-topic instance. The study identified 276 focal sets; the top two focal sets are focal sets 275 and 276. The most important focal content comes from an Indonesian Twitter user, who operates a personal blog on opinion and story covers. The findings highlight the effectiveness of multisource KG-CFSA in establishing context for a social network analysis.
In recent decades, several community detection methods have been proposed to uncover densely connected groups in a network. In many cases, there is the necessity of detecting such a group around a seed node that may be of particular topological importance or the external knowledge enable that node to be significant for the specific subgroup. In this way, local community detection algorithms can discover subgroups faster than global community detection algorithms that cover the entire network. This study aims to introduce new variants of an existing local community detection algorithm that uncover a single community and establish the positive aspects of identifying better communities. We experimentally evaluate our proposed methods in synthetic and real datasets compared to the baseline method, three state-of-the-art methods in local community detection, and a global community detection method.
In scenarios where objects are characterized by a combination of rigid and flexible features, we consider the problem of identifying a natural set of rigid clusters, along with a network model of flexible states per cluster. Our approach proves effective within the Allais paradox context. Our algorithm, applied to data collected in an experiment, identified personality clusters and emotion states within each cluster. This model outperforms alternative clustering models in capturing information regarding participants' choices. Beyond the current scope, our approach is applicable to other data-sets with combined rigid-flexible attributes. Beyond prediction, a strategy that aims to achieve a result by influencing a flexible state holds the promise of enhanced effectiveness when it is tailored to a cluster.
Identifying the influential spreaders in complex networks has emerged as an important research challenge to control the spread of (mis)information or infectious diseases. Researchers have proposed many centrality measures to identify the influential nodes (spreaders) in the past few years. Still, most of them have not considered the importance of the edges in unweighted networks. To address this issue, we propose a novel centrality measure to identify the spreading ability of the Influential Spreaders using the Potential Edge Weight method (IS-PEW). Considering the connectivity structure, the ability of information exchange, and the importance of neighbouring nodes, we measure the potential edge weight. The ranking similarity of spreaders identified by IS-PEW and the baseline centrality methods are compared with the Susceptible-Infectious-Recovered (SIR) epidemic simulator using Kendall’s rank correlation. The spreading ability of the top-ranking spreaders is also compared for five different percentages of top-ranking node sets using six different real networks.
