Local Subgraph Research Articles

GNNGL-PPI: multi-category prediction of protein-protein interactions using graph neural networks based on global graphs and local subgraphs

Most proteins exert their functions by interacting with other proteins, making the identification of protein-protein interactions (PPI) crucial for understanding biological activities, pathological mechanisms, and clinical therapies. Developing effective and reliable computational methods for predicting PPI can significantly reduce the time-consuming and labor-intensive associated traditional biological experiments. However, accurately identifying the specific categories of protein-protein interactions and improving the prediction accuracy of the computational methods remain dual challenges. To tackle these challenges, we proposed a novel graph neural network method called GNNGL-PPI for multi-category prediction of PPI based on global graphs and local subgraphs. GNNGL-PPI consisted of two main components: using Graph Isomorphism Network (GIN) to extract global graph features from PPI network graph, and employing GIN As Kernel (GIN-AK) to extract local subgraph features from the subgraphs of protein vertices. Additionally, considering the imbalanced distribution of samples in each category within the benchmark datasets, we introduced an Asymmetric Loss (ASL) function to further enhance the predictive performance of the method. Through evaluations on six benchmark test sets formed by three different dataset partitioning algorithms (Random, BFS, DFS), GNNGL-PPI outperformed the state-of-the-art multi-category prediction methods of PPI, as measured by the comprehensive performance evaluation metric F1-measure. Furthermore, interpretability analysis confirmed the effectiveness of GNNGL-PPI as a reliable multi-category prediction method for predicting protein-protein interactions.

Fast Local Subgraph Counting

Fast Local Subgraph Counting

DualGAD: Dual-bootstrapped self-supervised learning for graph anomaly detection

Graph anomaly detection (GAD) is an emerging and essential research field for discovering anomalous individuals (e.g., nodes or edges) that deviate significantly from the normal majority in an attributed graph. Unlike other anomaly types (e.g., image, text) with independence, it is not trivial to capture inherent and distinctive anomaly patterns as graph anomalies usually exist more complex relational interaction with a overwhelming amount of normal majority, which induces the inconsistency in feature and structure space. Recently, some studies have applied graph self-supervised learning on the GAD task and surge a new climax. Despite their success, the sub-optimal performance is only achieved because the imbalance nature of anomaly problem significantly dilutes anomalous information and causes the shortage of effective supervision signals. To address these challenges, we propose a dual-bootstrapped self-supervised approach, namely DualGAD, that consists of one generative module and one contrastive module. Specifically, we first sample a local subgraph for each target node and employ two mask strategies on subgraphs to break the short-range connection and reduce redundancy spread from majority neighbors. The generative module is equipped with the reconstruction objective to model the surrounding context of masked subgraphs, which learn discriminative representations by capturing the inconsistency patterns at both feature and structure space. Considering the local information is easily been over-emphasized, we elaborately tailor a novel cluster-guided contrastive learning module without relying on positive/negative pairs to learn the intrinsic anomaly property by aligning the perturbation distributions across views. Finally, the two self-supervised modules are seamlessly integrated and bootstrap each other for learning the discriminative representations to distinguish anomalies and normal majority. Extensive experiments demonstrate the superiority of DualGAD on the node-level anomaly detection task.

Towards Inductive Robustness: Distilling and Fostering Wave-Induced Resonance in Transductive GCNs against Graph Adversarial Attacks

Graph neural networks (GNNs) have recently been shown to be vulnerable to adversarial attacks, where slight perturbations in the graph structure can lead to erroneous predictions. However, current robust models for defending against such attacks inherit the transductive limitations of graph convolutional networks (GCNs). As a result, they are constrained by fixed structures and do not naturally generalize to unseen nodes. Here, we discover that transductive GCNs inherently possess a distillable robustness, achieved through a wave-induced resonance process. Based on this, we foster this resonance to facilitate inductive and robust learning. Specifically, we first prove that the signal formed by GCN-driven message passing (MP) is equivalent to the edge-based Laplacian wave, where, within a wave system, resonance can naturally emerge between the signal and its transmitting medium. This resonance provides inherent resistance to malicious perturbations inflicted on the signal system. We then prove that merely three MP iterations within GCNs can induce signal resonance between nodes and edges, manifesting as a coupling between nodes and their distillable surrounding local subgraph. Consequently, we present Graph Resonance-fostering Network (GRN) to foster this resonance via learning node representations from their distilled resonating subgraphs. By capturing the edge-transmitted signals within this subgraph and integrating them with the node signal, GRN embeds these combined signals into the central node's representation. This node-wise embedding approach allows for generalization to unseen nodes. We validate our theoretical findings with experiments, and demonstrate that GRN generalizes robustness to unseen nodes, whilst maintaining state-of-the-art classification accuracy on perturbed graphs. Appendices can be found on arXiv version: https://arxiv.org/abs/2312.08651

Fortune favors the invariant: Enhancing GNNs’ generalizability with Invariant Graph Learning

Generalizable and transferrable graph representation learning endows graph neural networks (GNN) with the ability to extrapolate potential test distributions. Nonetheless, current endeavors recklessly ascribe the demoralizing performance on a single entity (feature or edge) distribution shift and resort to uncontrollable augmentation. By inheriting the philosophy of Invariant graph learning (IGL), which characterizes a full graph as an invariant core subgraph (rationale) and a complementary trivial part (environment), we propose a universal operator termed InMvie to release GNN’s out-of-distribution generation potential. The advantages of our proposal can be attributed to two main factors: the comprehensive and customized insight on each local subgraph, and the systematical encapsulation of environmental interventions. Concretely, a rationale miner is designed to find a small subset of the input graph — rationale, which injects the model with feature invariance while filtering out the spurious patterns, i.e., environment. Then, we utilize systematic environment intervention to ensure the out-of-distribution awareness of the model. Our InMvie has been validated through experiments on both synthetic and real-world datasets, proving its superiority in terms of interpretability and generalization ability for node classification over the leading baselines.

Reconstructed Graph Neural Network With Knowledge Distillation for Lightweight Anomaly Detection.

The proliferation of Internet-of-Things (IoT) technologies in modern smart society enables massive data exchange for offering intelligent services. It becomes essential to ensure secure communications while exchanging highly sensitive IoT data efficiently, which leads to high demands for lightweight models or algorithms with limited computation capability provided by individual IoT devices. In this study, a graph representation learning model, which seamlessly incorporates graph neural network (GNN) and knowledge distillation (KD) techniques, named reconstructed graph with global-local distillation (RG-GLD), is designed to realize the lightweight anomaly detection across IoT communication networks. In particular, a new graph network reconstruction strategy, which treats data communications as nodes in a directed graph while edges are then connected according to two specifically defined rules, is devised and applied to facilitate the graph representation learning in secure and efficient IoT communications. Both the structural and traffic features are then extracted from the graph data and flow data respectively, based on the graph attention network (GAT) and multilayer perceptron (MLP) techniques. These can benefit the GNN-based KD process in accordance with the more effective feature fusion and representation, considering both structural and data levels across the dynamic IoT networks. Furthermore, a lightweight local subgraph preservation mechanism improved by the graph attention mechanism and downsampling scheme to better utilize the topological information, and a so-called global information alignment defined based on the self-attention mechanism to effectively preserve the global information, are developed and incorporated in a refined graph attention based KD scheme. Compared with four different baseline methods, experiments and evaluations conducted based on two public datasets demonstrate the usefulness and effectiveness of our proposed model in improving the efficiency of knowledge transfer with higher classification accuracy but lower computational load, which can be deployed for lightweight anomaly detection in sustainable IoT computing environments.

Balancing structure and position information in Graph Transformer network with a learnable node embedding

The Transformer-based graph neural network models have achieved remarkable results in graph representation learning in recent years. One of the main challenges in graph representation learning with Transformer architecture is the non-existence of a universal positional encoding. Standard position encoding methods usually evolve the usage of the graph Laplacian matrix eigenvectors. However, exploiting the structural information from these eigenvectors failed to perform graph learning tasks requiring the node’s local structures. In our work, we propose a novel node encoding that leverages both the node’s global position information and the node’s local structural information, which can generalize well for a wide range of graph learning tasks. The global position encoding branch operates on the eigenvalues and eigenvectors of the Laplacian matrix of the entire graph. The structural encoding branch is derived through the spectral-based encoding of the local subgraph. It represents the local properties, which are usually omitted in the Laplacian position encoding because of the cutoff of high graph frequencies. Two encoding branches are designed with learnable weights and mapped into predefined embedding spaces. Then, a weighted combination is employed to create a unique location encoding for each node. We validate the efficiency of our proposed encoding through various graph learning datasets, including node classification, link prediction, graph classification, and graph regression tasks. The overall results demonstrate that our structural and positional encoding can balance between the local and global structural information and outperforms most of the baseline models.

The Application of Micro-Video Technology in the Promotion of Digital Reading in Children's Libraries Under the Background of Media Fusion

Digital reading is now an important way for children to read. This article defines the concept of children's digital reading based on the current situation, reading needs, and research progress of reading promotion methods. Aiming at the problem that only simple matching method is used for entity extraction in BAG model, this paper proposes an entity extraction method based on reference words and an entity graph construction method based on problem-related entities. Experiments on data sets verify the effectiveness of the model. The results show that the BAG model is improved by 2.4% on the verification set and 4.1% on the test set, which proves the effectiveness of the improved method. Based on the graph convolution algorithm of the central node, the local subgraph convolution and the global graph convolution are connected in the entity graph behind the central node, so that the model can obtain local and global information at the same time and enhance the reasoning ability of the model.

An optimized task assignment framework based on crowdsourcing knowledge graph and prediction

In spatial crowdsourcing, both workers and tasks are added to the system on the fly, making it a real-time platform with complex circumstances. Most research has focused on optimizing the total number of assigned tasks, with little attention paid to factors such as worker preferences and task completion rate in a practical environment. To capture the complex spatiotemporal correlation and the latent supply–demand relationship, this work proposes a convolutional spatiotemporal attention model (Conv-STAN) and a cluster-based time-weighted voting method (CT-Voting) for predicting the future distribution of crowdsourcing entities. We develop a crowdsourcing knowledge graph (CKG) that reflects the semantic linkages between entities and provide an innovative and effective state representation approach (CKG2vec) for quantifying the preference to demonstrate the impact of a dynamically changing environment on worker preferences. In addition, a local subgraph incremental updating approach (LSG-IncrUpdate) is constructed to simulate the interaction between various entities in a crowdsourcing setup. Finally, we introduce a reinforcement learning framework (CKG-RLTA) that integrates prediction and the crowdsourcing knowledge graph to conduct task assignment. The experimental findings show that our approach is sensitive to workers’ preferences and improves task completion rate, even when those preferences are subject to change.

Exact tests for offline changepoint detection in multichannel binary and count data with application to networks

We consider offline detection of a single changepoint in binary and count time-series. We compare exact tests based on the cumulative sum (CUSUM) and the likelihood ratio (LR) statistics, and a new proposal that combines exact two-sample conditional tests with multiplicity correction, against standard asymptotic tests based on the Brownian bridge approximation to the CUSUM statistic. We see empirically that the exact tests are much more powerful in situations where normal approximations driving asymptotic tests are not trustworthy: (i) small sample settings; (ii) sparse parametric settings; (iii) time-series with changepoint near the boundary. We also compare these exact tests against their bootstrap-based variants which do not have exact type I error control guarantees. Further, we consider a multichannel version of the problem, where channels can have different changepoints. Controlling the False Discovery Rate (FDR), we simultaneously detect changes in multiple channels. This ‘local’ approach is shown to be more advantageous than multivariate global testing approaches when the number of channels with changepoints is much smaller than the total number of channels. As a natural application, we consider network-valued time-series and use our approach with (a) edges as binary channels and (b) node-degrees or other local subgraph statistics as count channels. The local testing approach is seen to be much more informative than global network changepoint algorithms.

Influential Incremental Learning-Based Privacy Preservation for Social Network Information

Social network influence dissemination focuses on employing a small number of seed sets to generate the most significant possible influence in social networks and considers forwarding to be the only technique of information transmission, ignoring all other ways. Users, for example, can post a message via this mode of distribution (called para), which is difficult to trace, posing a danger of privacy leakage. This research tries to address the aforementioned issues by developing a social network information transmission model that supports the paranormal relationship. It suggests a way of disseminating information called Local Greedy, which aids in the protection of user privacy. Its effect helps to reconcile the conflict between privacy protection and information distribution. Aiming at the enumeration problem of seed set selection, an incremental strategy that supports privacy protection is proposed to construct seed sets to reduce time overhead; a local influence subgraph method of computing nodes is given to estimate the influence of seed set propagation quickly; the group satisfies the constraints of privacy protection, and a plan is proposed to deduce the upper limit of the probability of node leakage state, avoiding the time cost of using the Monte Carlo method using the crawled Sina Weibo dataset. Experimental verification and example analysis are carried out, and the results show the effectiveness of the proposed method.

Two-geodesic-transitive graphs which are locally self-complementary

A 2-geodesic of a graph is a vertex triple (u,v,w) with v adjacent to both u and w, u≠w and u,w are not adjacent. A graph is said to be 2-geodesic-transitive if its automorphism group is transitive on both the set of arcs and the set of 2-geodesics. In this paper, we first determine the family of 2-geodesic-transitive graphs which are locally self-complementary, and then classify the family of 2-geodesic-transitive graphs that the local subgraph induced by the neighbor of a vertex is an arc-transitive circulant.

A localized distributed algorithm for vertex cover problem

Finding the minimum vertex cover of a given graph is a well-known NP-Hard problem that has many applications in various fields. In this paper, we propose a distributed localized algorithm for detecting vertex cover using 2-hop local neighborhood information in the distributed systems. We propose a scoring based policy and add the 1-hop neighbors of nodes with the highest score among their 2-hop neighbors to the vertex cover. The score of nodes is calculated by dividing the number of uncovered edges in their local subgraph by the number of their 1-hop neighbors. In this way, the score of each node determines the average coverage ratio by each neighbor of that node. The time and bit complexities of the proposed algorithm are O(n/Δ) and O(n2×logn) respectively, where Δ is the maximum node degree and n is the number of nodes. The comprehensive simulation results showed that the proposed algorithm could find up to 11% smaller solutions than the existing distributed algorithms with less than 1% difference of optimum solutions in most of the evaluated graphs.

Local clustering via approximate heat kernel PageRank with subgraph sampling

Graph clustering, a fundamental technique in network science for understanding structures in complex systems, presents inherent problems. Though studied extensively in the literature, graph clustering in large systems remains particularly challenging because massive graphs incur a prohibitively large computational load. The heat kernel PageRank provides a quantitative ranking of nodes, and a local cluster can be efficiently found by performing a sweep over the heat kernel PageRank vector. But computing an exact heat kernel PageRank vector may be expensive, and approximate algorithms are often used instead. Most approximate algorithms compute the heat kernel PageRank vector on the whole graph, and thus are dependent on global structures. In this paper, we present an algorithm for approximating the heat kernel PageRank on a local subgraph. Moreover, we show that the number of computations required by the proposed algorithm is sublinear in terms of the expected size of the local cluster of interest, and that it provides a good approximation of the heat kernel PageRank, with approximation errors bounded by a probabilistic guarantee. Numerical experiments verify that the local clustering algorithm using our approximate heat kernel PageRank achieves state-of-the-art performance.

Genome-scale de novo assembly using ALGA.

MotivationThere are very few methods for de novo genome assembly based on the overlap graph approach. It is considered as giving more exact results than the so-called de Bruijn graph approach but in much greater time and of much higher memory usage. It is not uncommon that assembly methods involving the overlap graph model are not able to successfully compute greater datasets, mainly due to memory limitation of a computer. This was the reason for developing in last decades mainly de Bruijn-based assembly methods, fast and fairly accurate. However, the latter methods can fail for longer or more repetitive genomes, as they decompose reads to shorter fragments and lose a part of information. An efficient assembler for processing big datasets and using the overlap graph model is still looked out.ResultsWe propose a new genome-scale de novo assembler based on the overlap graph approach, designed for short-read sequencing data. The method, ALGA, incorporates several new ideas resulting in more exact contigs produced in short time. Among these ideas, we have creation of a sparse but quite informative graph, reduction of the graph including a procedure referring to the problem of minimum spanning tree of a local subgraph, and graph traversal connected with simultaneous analysis of contigs stored so far. What is rare in genome assembly, the algorithm is almost parameter-free, with only one optional parameter to be set by a user. ALGA was compared with nine state-of-the-art assemblers in tests on genome-scale sequencing data obtained from real experiments on six organisms, differing in size, coverage, GC content and repetition rate. ALGA produced best results in the sense of overall quality of genome reconstruction, understood as a good balance between genome coverage, accuracy and length of resulting sequences. The algorithm is one of tools involved in processing data in currently realized national project Genomic Map of Poland.Availability and implementationALGA is available at http://alga.put.poznan.pl.Supplementary information Supplementary data are available at Bioinformatics online.

Distributed subgraph counting

In this paper, we study local subgraph counting, which is to count the occurrences of a user-given pattern graph p around every node v in a data graph G , when v matches to a given orbit o in p , where the orbit serves as a center to count p. In general, the orbit can be a node, an edge, or a set of nodes in p. Local subgraph counting has played an important role in characterizing high-order local structures that exhibit in a large graph, and has been widely used in denser and relevant communities mining, graphlet degree distribution, discriminative features selection for link prediction, relational classification and recommendation. In the literature, almost all the existing works support a k -node pattern graph, for k ≤ 5, with either 1 node orbit or 1 edge orbit. Their approaches are difficult to support larger k due to the fact that subgraph counting is to count by subgraph isomorphism. In this work, we develop a new general approach to count any k pattern graphs with any orbits selected. The key idea behind is that we do local subgraph counting by homomorphism counting, which can be solved by relational algebra using joins, group-by and aggregation. By homomorphism counting, we do local subgraph counting by eliminating counts for those that are not subgraph isomorphism matchings from the total count for any possible matchings. We have developed a distributed system named DISC on Spark. Our extensive experiments validate the efficiency of our approach by testing 114 local subgraph counting queries used in the existing work over real graphs, where no existing work can support all.

Efficient ($$\alpha $$, $$\beta $$)-core computation in bipartite graphs

The problem of computing ( $$\alpha , \beta $$ )-core in a bipartite graph for given $$\alpha $$ and $$\beta $$ is a fundamental problem in bipartite graph analysis and can be used in many applications such as online group recommendation and fraudsters detection Existing solution to computing ( $$\alpha , \beta $$ )-core needs to traverse the entire bipartite graph once and ignore the fact that real-world graphs are often dynamic. Considering the real bipartite graph can be very large and dynamically updated, and the requests to compute ( $$\alpha , \beta $$ )-core can be issued frequently in real applications, the existing solution is too expensive to compute the $$(\alpha ,\beta )$$ -core. In this paper, we present an efficient algorithm for ( $$\alpha , \beta $$ )-core computation based on a novel index such that the algorithm runs in linear time regarding the result size (thus, the algorithm is optimal since it needs at least linear time to output the result). We prove that the index only requires O(m) space where m is the number of edges in the bipartite graph. We also devise an efficient algorithm with time complexity $$O (\delta \cdot m)$$ for index construction where $$\delta $$ is bounded by $$\sqrt{m}$$ and is much smaller than $$\sqrt{m}$$ in practice. Moreover, we discuss efficient algorithms to maintain the index when the bipartite graph is dynamically updated. We show that we can decide whether a node in the index should be updated or not by visiting its neighbors. Based on this locality property, we propose an efficient index maintenance algorithm which only needs to visit a local subgraph near the inserted or removed edge. Finally, we show how to implement our index construction and maintenance algorithms in parallel. The experimental results on real and synthetic graphs (more than 1 billion edges) demonstrate that our algorithms achieve up to 5 orders of magnitude speedup for computing $$(\alpha ,\beta )$$ -core, up to 3 orders of magnitude speedup for index construction and up to 4 orders of magnitude speedup for index maintenance, respectively, compared with existing techniques.

Protein complex detection algorithm based on multiple topological characteristics in PPI networks

Detecting protein complexes from available protein–protein interaction (PPI) networks is an important task, and several related algorithms have been proposed. These algorithms usually consider a single topological metric and ignore the rich topological characteristics and inherent organization information of protein complexes. However, the effective use of such information is crucial to protein complex detection. To overcome this deficiency, this study presents a heuristic clustering algorithm to identify protein complexes by fully exploiting the topological information of PPI networks. By considering the clustering coefficient and the node degree, a new nodal metric is proposed to quantify the importance of each node within a local subgraph. An iterative paradigm is used to incrementally identify seed proteins and expand each seed to a cluster. First, among the unclustered nodes, the node with the highest nodal metric is selected as a new seed. Then, the seed is expanded to a cluster by adding candidate nodes recursively from its neighbors according to both the density of the cluster and the connection between a candidate node and the cluster. The experimental results demonstrate that the proposed algorithm outperforms other competing algorithms in terms of F-measure and accuracy.

Design and Evaluation of Algorithms for Energy Efficient and Complete Determination of Critical Nodes for Wireless Sensor Network Reliability

A critical node (cut vertex or articulation point) in wireless sensor networks, is a node which its failure breaks the connectivity of the network. Therefore, it is crucial that critical nodes be detected and treated with caution. This paper provides two localized distributed algorithms for determining the states of nodes (critical or noncritical). The first proposed algorithm identifies most of the critical and noncritical dominator nodes from two-hop local subgraph and connected dominating set (CDS) information that limits the computational complexity to $O(\Delta ^2)$ and bit complexity to $O(c\text{log}_2n)$ where $\Delta$ is the maximum node degree, $c$ is the critical node count, and $n$ is the node count. The testbed experiments and simulation results show that this algorithm detects up to 93% of critical nodes and achieves up to 91% of state determination with low energy consumption. The second proposed algorithm, which is based on the first one, finds the states of all nodes by running a limited distributed depth-first search algorithm in unrecognized parts of the network without traversing the whole network. Comprehensive testbed experiments and simulation results reveal that, in the presence of a CDS, this algorithm finds all critical nodes with lower energy consumption than all existing algorithms.

Evaluation of invulnerability of scale-free networks using total information of local sub-graph

The quantification for the invulnerability of complex network is a fundamental problem in which identifying influential nodes is of theoretical and practical significance. In this paper, we propose a novel definition of centrality named total information (TC) which derives from a local sub-graph being constructed by a node and its neighbors. The centrality is then defined as the sum of the self-information of the node and the mutual information of its neighbor nodes. We use the proposed centrality to identify the importance of nodes through the evaluation of the invulnerability of scale-free networks. It shows both the efficiency and the effectiveness of the proposed centrality are improved, compared with traditional centralities.