Inductive Representation Learning on Large Graphs

Hamilton, William L.; Ying, Rex; Leskovec, Jure

doi:10.48550/arxiv.1706.02216

Cited by 249 publications

(323 citation statements)

References 14 publications

Supporting

Mentioning

323

Contrasting

Order By: Relevance

“…This kind of GNNs only considers passing messages uniformly from one-or two-hop neighbors along edges. GraphSAGE [1] aggregates and updates features in a range of the two-hop neighbors to the center node. DGCN [8] considers the first-and second-order proximity to aggregate the attributes on the directed graphs.…”

Section: Graph Neural Networkmentioning

confidence: 99%

“…Node classification is a primary graph task for a wide range of applications [1,2,3,4,5,6], as it determines the category of the central node by comparing neighbors based on the messagepassing mechanism which learns the node attributes in the multiclass graph dataset. The distribution of multi categories in the graph dataset is directly tied to the network's overall performance.…”

Section: Graph Data Augmentationmentioning

confidence: 99%

“…Graph neural networks (GNNs) offer effective graph-based techniques applied to solve abundant real-world problems in diverse fields, such as social science [1], physical systems [2,3], protein-protein interaction networks [4], brain neuroscience [5], knowledge graphs [6], etc. The power of current GNNs [1,7,8,9,10] is largely due to their message-passing mechanisms. However, the underlying behavior that spontaneously aggregates messages on the graph structure is obscure.…”

Section: Introductionmentioning

confidence: 99%

“…In some research [1,7,8], messages were passed along edges uniformly without accounting for priority of either graph structure or node attributes. Intuitively, each neighbor node's impact was distinctive to the center node in the node classification task.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Graph Decipher: A transparent dual-attention graph neural network to understand the message-passing mechanism for the node classification

Yan¹,

Liu²

2022

Preprint

View full text Add to dashboard Cite

Graph neural networks can be effectively applied to find solutions for many real-world problems across widely diverse fields. The success of graph neural networks is linked to the message-passing mechanism on the graph, however the message-aggregating behavior is still not entirely clear in most algorithms. To improve functionality, we propose a new transparent network called Graph Decipher to investigate the message-passing mechanism by prioritizing in two main components: the graph structure and node attributes, at the graph, feature, and global levels on a graph under the node classification task. However the computation burden now becomes the most significant issue because the relevance of both graph structure and node attributes are computed on a graph. In order to solve this issue, only relevant representative node attributes are extracted by graph feature filters, allowing calculations to be performed in a category-oriented manner. Experiments on seven datasets show that Graph Decipher achieves state-of-the-art performance while imposing a substantially lower computation burden under the node classification task. Additionally, since our algorithm has the ability to explore the representative node attributes by category, it is utilized to alleviate the imbalanced node classification problem on multi-class graph datasets.

show abstract

Section: Graph Neural Networkmentioning

confidence: 99%

Section: Graph Data Augmentationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Graph Decipher: A transparent dual-attention graph neural network to understand the message-passing mechanism for the node classification

Yan¹,

Liu²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…This has been mainly addressed by pretrained generalizable models to infer on new disasters [11,12,13]. While GCNs initially were unable to be used on unseen contexts [7], the SAGE (sample and aggregate) framework has been proposed to learn inductively on mini-batches of graphs and can be used to predict on unseen graphs [14].…”

Section: Introductionmentioning

confidence: 99%

Towards Cross-Disaster Building Damage Assessment with Graph Convolutional Networks

Ismail¹,

Awad²

2022

Preprint

View full text Add to dashboard Cite

In the aftermath of disasters, building damage maps are obtained using change detection to plan rescue operations. Current convolutional neural network approaches do not consider the similarities between neighboring buildings for predicting the damage. We present a novel graph-based building damage detection solution to capture these relationships. Our proposed model architecture learns from both local and neighborhood features to predict building damage. Specifically, we adopt the sample and aggregate graph convolution strategy to learn aggregation functions that generalize to unseen graphs which is essential for alleviating the time needed to obtain predictions for new disasters. Our experiments on the xBD dataset and comparisons with a classical convolutional neural network reveal that while our approach is handicapped by class imbalance, it presents a promising and distinct advantage when it comes to cross-disaster generalization.

show abstract

Cuproptosis gene‐related, neural network‐based prognosis prediction and drug‐target prediction for KIRC

Liu,

Shao,

Hao

et al. 2023

Cancer Medicine

View full text Add to dashboard Cite

BackgroundKidney renal clear cell carcinoma (KIRC), as a common case in renal cell carcinoma (RCC), has the risk of postoperative recurrence, thus its prognosis is poor and its prognostic markers are usually based on imaging methods, which have the problem of low specificity. In addition, cuproptosis, as a novel mode of cell death, has been used as a biomarker to predict disease in many cancers in recent years, which also provides an important basis for prognostic prediction in KIRC. For postoperative patients with KIRC, an important means of preventing disease recurrence is pharmacological treatment, and thus matching the appropriate drug to the specific patient's target is also particularly important. With the development of neural networks, their predictive performance in the field of medical big data has surpassed that of traditional methods, and this also applies to the field of prognosis prediction and drug‐target prediction.ObjectiveThe purpose of this study is to screen for cuproptosis genes related to the prognosis of KIRC and to establish a deep neural network (DNN) model for patient risk prediction, while also developing a personalized nomogram model for predicting patient survival. In addition, sensitivity drugs for KIRC were screened, and a graph neural network (GNN) model was established to predict the targets of the drugs, in order to discover potential drug action sites and provide new treatment ideas for KIRC.MethodsWe used the Cancer Genome Atlas (TCGA) database, International Cancer Genome Consortium (ICGC) database, and DrugBank database for our study. Differentially expressed genes (DEGs) were screened using TCGA data, and then a DNN‐based risk prediction model was built and validated using ICGC data. Subsequently, the differences between high‐ and low‐risk groups were analyzed and KIRC‐sensitive drugs were screened, and finally a GNN model was trained using DrugBank data to predict the relevant targets of these drugs.ResultsA prognostic model was built by screening 10 significantly different cuproptosis‐related genes, the model had an AUC of 0.739 on the training set (TCGA data) and an AUC of 0.707 on the validation set (ICGC data), which demonstrated a good predictive performance. Based on the prognostic model in this paper, patients were also classified into high‐ and low‐risk groups, and functional analyses were performed. In addition, 251 drugs were screened for sensitivity, and four drugs were ultimately found to have high sensitivity, with 5‐Fluorouracil having the best inhibitory effect, and subsequently their corresponding targets were also predicted by GraphSAGE, with the most prominent targets including Cytochrome P450 2D6, UDP‐glucuronosyltransferase 1A, and Proto‐oncogene tyrosine‐protein kinase receptor Ret. Notably, the average accuracy of GraphSAGE was 0.817 ± 0.013, which was higher than that of GAT and GTN.ConclusionOur KIRC risk prediction model, constructed using 10 cuproptosis‐related genes, had good independent prognostic ability. In addition, we screened four highly sensitive drugs and predicted relevant targets for these four drugs that might treat KIRC. Finally, literature research revealed that four drug‐target interactions have been demonstrated in previous studies and the remaining targets are potential sites of drug action for future research.

show abstract

Inductive Representation Learning on Large Graphs

Cited by 249 publications

References 14 publications

Graph Decipher: A transparent dual-attention graph neural network to understand the message-passing mechanism for the node classification

Graph Decipher: A transparent dual-attention graph neural network to understand the message-passing mechanism for the node classification

Towards Cross-Disaster Building Damage Assessment with Graph Convolutional Networks

Cuproptosis gene‐related, neural network‐based prognosis prediction and drug‐target prediction for KIRC

Contact Info

Product

Resources

About