Deep Graph Convolutional Network Research Articles

Structure-aware protein-protein interaction site prediction using deep graph convolutional network.

Protein-protein interactions (PPI) play crucial roles in many biological processes, and identifying PPI sites is an important step for mechanistic understanding of diseases and design of novel drugs. Since experimental approaches for PPI site identification are expensive and time-consuming, many computational methods have been developed as screening tools. However, these methods are mostly based on neighbored features in sequence, and thus limited to capture spatial information. We propose a deep graph-based framework deep Graph convolutional network for Protein-Protein-Interacting Site prediction (GraphPPIS) for PPI site prediction, where the PPI site prediction problem was converted into a graph node classification task and solved by deep learning using the initial residual and identity mapping techniques. We showed that a deeper architecture (up to eight layers) allows significant performance improvement over other sequence-based and structure-based methods by more than 12.5% and 10.5% on AUPRC and MCC, respectively. Further analyses indicated that the predicted interacting sites by GraphPPIS are more spatially clustered and closer to the native ones even when false-positive predictions are made. The results highlight the importance of capturing spatially neighboring residues for interacting site prediction. The datasets, the pre-computed features, and the source codes along with the pre-trained models of GraphPPIS are available at https://github.com/biomed-AI/GraphPPIS. The GraphPPIS web server is freely available at https://biomed.nscc-gz.cn/apps/GraphPPIS. Supplementary data are available at Bioinformatics online.

DeepGCNs-Att for Point Cloud Semantic Segmentation

Compared with traditional CNNs, Graph Convolutional Networks (GCNs), with the graph’s neural network structure, can better characterize non-Euclidean space. Furthermore, with the number of the network layers increasing, deep GCNs demonstrate outstanding performance in mining the partial relationship between the point cloud’s local features. However, the current deep GCNs algorithm cannot sufficiently exploit the point cloud’s global characteristics for semantic segmentation. This paper proposes a novel network structure called DeepGCNs-Att to efficiently aggregate global context features. Moreover, to speed up the computation, we add an Attention layer after the GCN Backbone Block to mutually enhance the connection between the distant points of the non-Euclidean space. Our model is tested on the standard benchmark S3DIS. By comparing with other deep GCNs algorithms, our DeepGCNs-Att’s mIoU has at least two per cent higher than that of other models and even shows excellent results in space complexity and computational complexity under the same number of GCN layers.

Unifying Node Labels, Features, and Distances for Deep Network Completion.

Collected network data are often incomplete, with both missing nodes and missing edges. Thus, network completion that infers the unobserved part of the network is essential for downstream tasks. Despite the emerging literature related to network recovery, the potential information has not been effectively exploited. In this paper, we propose a novel unified deep graph convolutional network that infers missing edges by leveraging node labels, features, and distances. Specifically, we first construct an estimated network topology for the unobserved part using node labels, then jointly refine the network topology and learn the edge likelihood with node labels, node features and distances. Extensive experiments using several real-world datasets show the superiority of our method compared with the state-of-the-art approaches.

NPU RGBD Dataset and a Feature-Enhanced LSTM-DGCN Method for Action Recognition of Basketball Players+

Computer vision-based action recognition of basketball players in basketball training and competition has gradually become a research hotspot. However, owing to the complex technical action, diverse background, and limb occlusion, it remains a challenging task without effective solutions or public dataset benchmarks. In this study, we defined 32 kinds of atomic actions covering most of the complex actions for basketball players and built the dataset NPU RGB+D (a large scale dataset of basketball action recognition with RGB image data and Depth data captured in Northwestern Polytechnical University) for 12 kinds of actions of 10 professional basketball players with 2169 RGB+D videos and 75 thousand frames, including RGB frame sequences, depth maps, and skeleton coordinates. Through extracting the spatial features of the distances and angles between the joint points of basketball players, we created a new feature-enhanced skeleton-based method called LSTM-DGCN for basketball player action recognition based on the deep graph convolutional network (DGCN) and long short-term memory (LSTM) methods. Many advanced action recognition methods were evaluated on our dataset and compared with our proposed method. The experimental results show that the NPU RGB+D dataset is very competitive with the current action recognition algorithms and that our LSTM-DGCN outperforms the state-of-the-art action recognition methods in various evaluation criteria on our dataset. Our action classifications and this NPU RGB+D dataset are valuable for basketball player action recognition techniques. The feature-enhanced LSTM-DGCN has a more accurate action recognition effect, which improves the motion expression ability of the skeleton data.

DeepGCNs: Making GCNs Go as Deep as CNNs.

Convolutional neural networks (CNNs) have been very successful at solving a variety of computer vision tasks such as object classification and detection, semantic segmentation, activity understanding, to name just a few. One key enabling factor for their great performance has been the ability to train very deep networks. Despite their huge success in many tasks, CNNs do not work well with non-euclidean data, which is prevalent in many real-world applications. Graph Convolutional Networks (GCNs) offer an alternative that allows for non-Eucledian data input to a neural network. While GCNs already achieve encouraging results, they are currently limited to architectures with a relatively small number of layers, primarily due to vanishing gradients during training. This work transfers concepts such as residual/dense connections and dilated convolutions from CNNs to GCNs in order to successfully train very deep GCNs. We show the benefit of using deep GCNs (with as many as 112 layers) experimentally across various datasets and tasks. Specifically, we achieve very promising performance in part segmentation and semantic segmentation on point clouds and in node classification of protein functions across biological protein-protein interaction (PPI) graphs. We believe that the insights in this work will open avenues for future research on GCNs and their application to further tasks not explored in this paper. The source code for this work is available at https://github.com/lightaime/deep_gcns_torch and https://github.com/lightaime/deep_gcns for PyTorch and TensorFlow implementations respectively.

Hyperspectral Image Classification Based on Deep Attention Graph Convolutional Network

Hyperspectral images (HSIs) have gained high spectral resolution due to recent advances in spectral imaging technologies. This incurs problems, such as an increased data scale and an increased number of bands for HSIs, which results in a complex correlation between different bands. In the applications of remote sensing and earth observation, ground objects represented by each HSI pixel are composed of physical and chemical non-Euclidean structures, and HSI classification (HIC) is becoming a more challenging task. To solve the above problems, we propose a framework based on a deep attention graph convolutional network (DAGCN). Specifically, we first integrate an attention mechanism into the spectral similarity measurement to aggregate similar spectra. Therefore, we propose a new similarity measurement method, i.e., the mixed measurement of a kernel spectral angle mapper and spectral information divergence (KSAM-SID), to aggregate similar spectra. Considering the non-Euclidean structural characteristics of HSIs, we design deep graph convolutional networks (DeepGCNs) as a feature extraction method to extract deep abstract features and explore the internal relationship between HSI data. Finally, we dynamically update the attention graph adjacency matrix to adapt to the changes in each feature graph. Experiments on three standard HSI data sets, namely, the Indian Pines, Pavia University, and Salinas data sets, demonstrate that the DAGCN outperforms the baselines in terms of various evaluation criteria. For example, on the Indian Pines data set, the overall accuracy of the proposed method achieves 98.61% when the training sample is 10%.

TAGCN: Station-level demand prediction for bike-sharing system via a temporal attention graph convolution network

Nowadays, bike-sharing is available in many cities, solving the problem of the last mile, and it is an environmental-friendly way to commute. However, there is a tidal phenomenon in the bike-sharing system, and the rents/returns of bikes at different stations are unbalanced. Thus, bikes at different stations need to be rebalanced regularly and station-level demand prediction plays an essential role in bike-sharing rebalancing. In this paper, a novel deep graph convolutional network (GCN) model with temporal attention (TAGCN) is proposed for bike check-out/in number prediction of each station. TAGCN can not only model the spatial and temporal dependency between varying stations, but also reflect the influence of different time granularity, which are hour-level, day-level and week-level time periodicity. With the help of well-designed temporal attention mechanism, our model can capture the dynamical temporal correlations and comprehensive spatial patterns in bike check-out/in flow effectively. The proposed model consistently outperforms state-of-the-art methods on four real-world bike-sharing datasets that are four seasons data of Divvy Bike System in Chicago.

Affordance-Based Grasping Point Detection Using Graph Convolutional Networks for Industrial Bin-Picking Applications.

Grasping point detection has traditionally been a core robotic and computer vision problem. In recent years, deep learning based methods have been widely used to predict grasping points, and have shown strong generalization capabilities under uncertainty. Particularly, approaches that aim at predicting object affordances without relying on the object identity, have obtained promising results in random bin-picking applications. However, most of them rely on RGB/RGB-D images, and it is not clear up to what extent 3D spatial information is used. Graph Convolutional Networks (GCNs) have been successfully used for object classification and scene segmentation in point clouds, and also to predict grasping points in simple laboratory experimentation. In the present proposal, we adapted the Deep Graph Convolutional Network model with the intuition that learning from n-dimensional point clouds would lead to a performance boost to predict object affordances. To the best of our knowledge, this is the first time that GCNs are applied to predict affordances for suction and gripper end effectors in an industrial bin-picking environment. Additionally, we designed a bin-picking oriented data preprocessing pipeline which contributes to ease the learning process and to create a flexible solution for any bin-picking application. To train our models, we created a highly accurate RGB-D/3D dataset which is openly available on demand. Finally, we benchmarked our method against a 2D Fully Convolutional Network based method, improving the top-1 precision score by 1.8% and 1.7% for suction and gripper respectively.

Android Malware Detection Based on Structural Features of the Function Call Graph

The openness of Android operating system not only brings convenience to users, but also leads to the attack threat from a large number of malicious applications (apps). Thus malware detection has become the research focus in the field of mobile security. In order to solve the problem of more coarse-grained feature selection and larger feature loss of graph structure existing in the current detection methods, we put forward a method named DGCNDroid for Android malware detection, which is based on the deep graph convolutional network. Our method starts by generating a function call graph for the decompiled Android application. Then the function call subgraph containing the sensitive application programming interface (API) is extracted. Finally, the function call subgraphs with structural features are trained as the input of the deep graph convolutional network. Thus the detection and classification of malicious apps can be realized. Through experimentation on a dataset containing 11,120 Android apps, the method proposed in this paper can achieve detection accuracy of 98.2%, which is higher than other existing detection methods.

Fault Diagnosis of Wind Turbine Gearbox Using a Novel Method of Fast Deep Graph Convolutional Networks

The fault diagnosis of the gearbox of wind turbines is a crucial task for wind turbine operation and maintenance. Although a convolutional neural network can extract the related information of adjacent sampling points using kernels, traditional deep learning methods have not leveraged related information from points with a large span of vibration signal data. In this article, a novel fast deep graph convolutional network is proposed to diagnose faults in the gearbox of wind turbines. First, the original vibration signals of the wind turbine gearbox are decomposed by wavelet packet, which presents time–frequency features as graphs. Then, graph convolutional networks are introduced to extract the features of points with a large span of the defined graph samples. Finally, the fast graph convolutional kernel and the particular pooling improvement are used to reduce the number of nodes and achieve fast classification. Experiments on two data sets are performed to verify the efficacy of the proposed method.

IDDkin: network-based influence deep diffusion model for enhancing prediction of kinase inhibitors.

Protein kinases have been the focus of drug discovery research for many years because they play a causal role in many human diseases. Understanding the binding profile of kinase inhibitors is a prerequisite for drug discovery, and traditional methods of predicting kinase inhibitors are time-consuming and inefficient. Calculation-based predictive methods provide a relatively low-cost and high-efficiency approach to the rapid development and effective understanding of the binding profile of kinase inhibitors. Particularly, the continuous improvement of network pharmacology methods provides unprecedented opportunities for drug discovery, network-based computational methods could be employed to aggregate the effective information from heterogeneous sources, which have become a new way for predicting the binding profile of kinase inhibitors. In this study, we proposed a network-based influence deep diffusion model, named IDDkin, for enhancing the prediction of kinase inhibitors. IDDkin uses deep graph convolutional networks, graph attention networks and adaptive weighting methods to diffuse the effective information of heterogeneous networks. The updated kinase and compound representations are used to predict potential compound-kinase pairs. The experimental results show that the performance of IDDkin is superior to the comparison methods, including the state-of-the-art kinase inhibitor prediction method and the classic model widely used in relationship prediction. In experiments conducted to verify its generalizability and in case studies, the IDDkin model also shows excellent performance. All of these results demonstrate the powerful predictive ability of the IDDkin model in the field of kinase inhibitors. Source code and data can be downloaded from https://github.com/CS-BIO/IDDkin. Supplementary data are available at Bioinformatics online.

Predicting functions of maize proteins using graph convolutional network

BackgroundMaize (Zea mays ssp. mays L.) is the most widely grown and yield crop in the world, as well as an important model organism for fundamental research of the function of genes. The functions of Maize proteins are annotated using the Gene Ontology (GO), which has more than 40000 terms and organizes GO terms in a direct acyclic graph (DAG). It is a huge challenge to accurately annotate relevant GO terms to a Maize protein from such a large number of candidate GO terms. Some deep learning models have been proposed to predict the protein function, but the effectiveness of these approaches is unsatisfactory. One major reason is that they inadequately utilize the GO hierarchy.ResultsTo use the knowledge encoded in the GO hierarchy, we propose a deep Graph Convolutional Network (GCN) based model (DeepGOA) to predict GO annotations of proteins. DeepGOA firstly quantifies the correlations (or edges) between GO terms and updates the edge weights of the DAG by leveraging GO annotations and hierarchy, then learns the semantic representation and latent inter-relations of GO terms in the way by applying GCN on the updated DAG. Meanwhile, Convolutional Neural Network (CNN) is used to learn the feature representation of amino acid sequences with respect to the semantic representations. After that, DeepGOA computes the dot product of the two representations, which enable to train the whole network end-to-end coherently. Extensive experiments show that DeepGOA can effectively integrate GO structural information and amino acid information, and then annotates proteins accurately.ConclusionsExperiments on Maize PH207 inbred line and Human protein sequence dataset show that DeepGOA outperforms the state-of-the-art deep learning based methods. The ablation study proves that GCN can employ the knowledge of GO and boost the performance. Codes and datasets are available at http://mlda.swu.edu.cn/codes.php?name=DeepGOA.

Efficient and Stable Graph Scattering Transforms via Pruning.

Graph convolutional networks (GCNs) have well-documented performance in various graph learning tasks, but their analysis is still at its infancy. Graph scattering transforms (GSTs) offer training-free deep GCN models that extract features from graph data, and are amenable to generalization and stability analyses. The price paid by GSTs is exponential complexity in space and time that increases with the number of layers. This discourages deployment of GSTs when a deep architecture is needed. The present work addresses the complexity limitation of GSTs by introducing an efficient so-termed pruned (p)GST approach. The resultant pruning algorithm is guided by a graph-spectrum-inspired criterion, and retains informative scattering features on-the-fly while bypassing the exponential complexity associated with GSTs. Stability of the novel pGSTs is also established when the input graph data or the network structure are perturbed. Furthermore, the sensitivity of pGST to random and localized signal perturbations is investigated analytically and experimentally. Numerical tests showcase that pGST performs comparably to the baseline GST at considerable computational savings. Furthermore, pGST achieves comparable performance to state-of-the-art GCNs in graph and 3D point cloud classification tasks. Upon analyzing the pGST pruning patterns, it is shown that graph data in different domains call for different network architectures, and that the pruning algorithm may be employed to guide the design choices for contemporary GCNs.

System-level prognostics and health management: A graph convolutional network–based framework

Sensing technologies have been used to gather massive amounts of data to improve system reliability analysis with the use of deep learning. Their use has been mainly focused on specific components or for the whole system, resulting in a drawback when dealing with complex systems as the interactions among components are not explicitly taken into account. Here, we propose a system-level prognostics and health management framework based on geometrical deep learning where a system, its components with their interactions, and sensor data are represented as a graph. This enables reliability analysis at different hierarchical levels by means of (1) a system-level module for system health diagnosis and prognosis based on embeddings of the system’s learned features from a graph convolutional network; (2) a component-level module based on a deep graph convolutional network for health state diagnosis for the system’s components; (3) a component interactions module based on a graph convolutional network autoencoder that allows for the identification of interactions among components when the system is in a degraded state. The framework is exemplified via a case study involving a chlorine dioxide generation system, in which it is shown that integrating both components’ interactions and sensor data in the form of a graph improves health state diagnosis capabilities.

Going Deep: Graph Convolutional Ladder-Shape Networks

Neighborhood aggregation algorithms like spectral graph convolutional networks (GCNs) formulate graph convolutions as a symmetric Laplacian smoothing operation to aggregate the feature information of one node with that of its neighbors. While they have achieved great success in semi-supervised node classification on graphs, current approaches suffer from the over-smoothing problem when the depth of the neural networks increases, which always leads to a noticeable degradation of performance. To solve this problem, we present graph convolutional ladder-shape networks (GCLN), a novel graph neural network architecture that transmits messages from shallow layers to deeper layers to overcome the over-smoothing problem and dramatically extend the scale of the neural networks with improved performance. We have validated the effectiveness of proposed GCLN at a node-wise level with a semi-supervised task (node classification) and an unsupervised task (node clustering), and at a graph-wise level with graph classification by applying a differentiable pooling operation. The proposed GCLN outperforms original GCNs, deep GCNs and other state-of-the-art GCN-based models for all three tasks, which were designed from various perspectives on six real-world benchmark data sets.

Intelligent acoustic-based fault diagnosis of roller bearings using a deep graph convolutional network

Roller bearings form key components in many machines and, as such, their health status can directly influence the operation of the entire machine. Acoustic signals collected from roller bearings contain information on their health status. Hence, acoustic-based fault diagnosis techniques can provide novel solutions as condition monitoring tools for roller bearings. Traditionally, acoustic fault diagnosis methods have been based on conventional signal processing methods in which prior expert knowledge has been required in order to extract and interpret the health information contained within the collected acoustic signals. As an alternative, deep learning methods can be used to obtain heath information from the collected signals by constructing ‘end-to-end’ models that do not rely on prior knowledge. These approaches have been successfully applied in the condition monitoring of industrial machinery. However, conventional deep learning methods can only learn features from the vertices of input data and thereby ignore the information contained in the relationships (edges) between vertices. In this paper, which combines graph convolution operators, graph coarsening methods, and graph pooling operations; a deep graph convolutional network (DGCN) based on graph theory is applied to deliver acoustic-based fault diagnosis of roller bearings. In the proposed method, the collected acoustic signals are first transformed into graphs with geometric structures. The edge weights represent the similarity between connected vertices, which enriches the input information and hence improves the classification accuracy of the deep learning methods applied. To verify the effectiveness of the proposed system, experiments with roller bearings of varying condition were carried out in the laboratory. The experimental results demonstrate that the DGCN method can be used to detect different kinds and severities of faults in roller bearings by learning from the constructed graphs. The results have been compared to those obtained using other, conventional, deep learning methods applied to the same datasets. These comparative tests demonstrate improved classification accuracy when using the DGCN method.

Bot prediction on social networks of Twitter in altmetrics using deep graph convolutional networks

In the context of smart cities, it is crucial to filter out falsified information spread on social media channels through paid campaigns or bot-user accounts that significantly influence communication networks across the social communities and may affect smart decision-making by the citizens. In this paper, we focus on two major aspects of the Twitter social network associated with altmetrics: (a) to analyze the properties of bots on Twitter networks and (b) to distinguish between bots and human accounts. Firstly, we employed state-of-the-art social network analysis techniques that exploit Twitter’s social network properties in novel altmetrics data. We found that 87% of tweets are affected by bots that are involved in the network’s dominant communities. We also found that, to some extent, community size and the degree of distribution in Twitter’s altmetrics network follow a power-law distribution. Furthermore, we applied a deep learning model, graph convolutional networks, to distinguish between organic (human) and bot Twitter accounts. The deployed model achieved the promising results, providing up to 71% classification accuracy over 200 epochs. Overall, the study concludes that bot presence in altmetrics-associated social media platforms can artificially inflate the number of social usage counts. As a result, special attention is required to eliminate such discrepancies when using altmetrics data for smart decision-making, such as research assessment either independently or complementary along with traditional bibliometric indices.

Fault Location in Power Distribution Systems via Deep Graph Convolutional Networks

This paper develops a novel graph convolutional network (GCN) framework for fault location in power distribution networks. The proposed approach integrates multiple measurements at different buses while taking system topology into account. The effectiveness of the GCN model is corroborated by the IEEE 123 bus benchmark system. Simulation results show that the GCN model significantly outperforms other widely-used machine learning schemes with very high fault location accuracy. In addition, the proposed approach is robust to measurement noise and data loss errors. Data visualization results of two competing neural networks are presented to explore the mechanism of GCN's superior performance. A data augmentation procedure is proposed to increase the robustness of the model under various levels of noise and data loss errors. Further experiments show that the model can adapt to topology changes of distribution networks and perform well with a limited number of measured buses.

Graph convolutional network for multi-label VHR remote sensing scene recognition

We address the problem of multi-label scene classification from Very High Resolution (VHR) satellite remote sensing (RS) images in this paper by exploring the deep graph convolutional network (GCN). Since a given VHR RS scene contains several local features, the traditional single-label classification frameworks do not convey the true semantics of the scene. The multi-label classification approaches, on the other hand, is expected to aid in better characterization of the area under consideration. Under the multi-label setup, we find it intuitive to represent a given image as a region adjacency graph (RAG) of the respective local regions. In order to extract discriminative features from such irregular structures for enhanced classification, we model the subsequent supervised learning problem in terms of a novel multi-label deep GCN. We validate the efficacy of the proposed technique on the popular UC-Merced dataset where our framework outperforms with respect to the literature.