Text Graphs Research Articles

An effective multi-modal adaptive contextual feature information fusion method for Chinese long text classification

Chinese long text classification plays a vital role in Natural Language Processing. Compared to Chinese short texts, Chinese long texts contain more complex semantic feature information. Furthermore, the distribution of these semantic features is uneven due to the varying lengths of the texts. Current research on Chinese long text classification models primarily focuses on enhancing text semantic features and representing Chinese long texts as graph-structured data. Nonetheless, these methods are still susceptible to noise information and tend to overlook the deep semantic information in long texts. To address the above challenges, this study proposes a novel and effective method called MACFM, which introduces a deep feature information mining method and an adaptive modal feature information fusion strategy to learn the semantic features of Chinese long texts thoroughly. First, we present the DCAM module to capture complex semantic features in Chinese long texts, allowing the model to learn detailed high-level representation features. Then, we explore the relationships between word vectors and text graphs, enabling the model to capture abundant semantic information and text positional information from the graph. Finally, we develop the AMFM module to effectively combine different modal feature representations and eliminate the unrelated noise information. The experimental results on five Chinese long text datasets show that our method significantly improves the accuracy of Chinese long text classification tasks. Furthermore, the generalization experiments on five English datasets and the visualized results demonstrate the effectiveness and interpretability of the MACFM model.

Graph-based Text Classification by Contrastive Learning with Text-level Graph Augmentation

Text Classification (TC) is a fundamental task in the information retrieval community. Nowadays, the mainstay TC methods are built on the deep neural networks, which can learn much more discriminative text features than the traditional shallow learning methods. Among existing deep TC methods, the ones based on Graph Neural Network (GNN) have attracted more attention due to the superior performance. Technically, the GNN-based TC methods mainly transform the full training dataset to a graph of texts; however, they often neglect the dependency between words, so as to miss potential semantic information of texts, which may be significant to exactly represent them. To solve the aforementioned problem, we generate graphs of words instead, so as to capture the dependency information of words. Specifically, each text is translated into a graph of words, where neighboring words are linked. We learn the node features of words by a GNN-like procedure and then aggregate them as the graph feature to represent the current text. To further improve the text representations, we suggest a contrastive learning regularization term. Specifically, we generate two augmented text graphs for each original text graph, we constrain the representations of the two augmented graphs from the same text close and the ones from different texts far away. We propose various techniques to generate the augmented graphs. Upon those ideas, we develop a novel deep TC model, namely Text-level Graph Networks with Contrastive Learning (TGN cl ). We conduct a number of experiments to evaluate the proposed TGN cl model. The empirical results demonstrate that TGN cl can outperform the existing state-of-the-art TC models.

A psychological evaluation method incorporating noisy label correction mechanism

Using machine learning and deep learning methods to analyze text data from social media can effectively explore hidden emotional tendencies and evaluate the psychological state of social media account owners. However, the label noise caused by mislabeling may significantly influence the training and prediction results of traditional supervised models. To resolve this problem, this paper proposes a psychological evaluation method that incorporates a noisy label correction mechanism and designs an evaluation framework that consists of a primary classification model and a noisy label correction mechanism. Firstly, the social media text data are transformed into heterogeneous text graphs, and a classification model combining a pre-trained model with a graph neural network is constructed to extract semantic features and structural features, respectively. After that, the Gaussian mixture model is used to select the samples that are likely to be mislabeled. Then, soft labels are generated for them to enable noisy label correction without prior knowledge of the noise distribution information. Finally, the corrected and clean samples are composed into a new data set and re-input into the primary model for mental state classification. Results of experiments on three real data sets indicate that the proposed method outperforms current advanced models in classification accuracy and noise robustness under different noise ratio settings, and can efficiently explore the potential sentiment tendencies and users’ psychological states in social media text data.

Sparse graph matching network for temporal language localization in videos

Temporal language localization in videos aims to retrieve the moment that best matches the text description in the untrimmed video using the query text. Existing methods using graph convolutional networks have been effective in feature representation and cross-modal interaction, but the existing methods do not consider the sparsity constraint of the graph when constructing the graph structure, which can easily cause an increase in computational cost and introduce redundant connections that may adversely affect the accuracy of the results. Therefore, we propose a novel sparse graph matching network for temporal language localization in videos. Specifically, we use graph convolutional networks to learn video features and dynamically construct video graph with constraints on sparsity and connectivity; the complementarity between the sequential context and the syntactic structure of text is used to model the semantic features of text. For cross-modal interaction, we design a sparse graph matching method based on affinity matrix to match video and text graphs, and align cross-modal semantic features. Finally, by fusing the features of the two modalities, candidate moments are generated, and their confidence scores are calculated to locate the moment matching the query. Experimental results on the public datasets TACoS, ActivityNet Caption, and QVHighlights demonstrate the superiority of our method compared to state-of-the-art methods.

An Effective Knowledgeable Label-Aware Approach for Sentential Relation Extraction

In recent years, sentential relation extraction has made remarkable progress with text and knowledge graphs (KGs). However, existing architectures ignore the valuable information contained in relationship labels, which KGs provide and can complement the model with additional signals and prior knowledge. To address this limitation, we propose a neural architecture that leverages knowledgeable labels to enhance sentential relation extraction. We name our proposed method knowledge label-aware sensitive relation extraction (KLA-SRE). To achieve this, we combine pre-trained static knowledge graph embeddings with learned semantic embeddings from other tokens to efficiently represent relation labels. By combining static pre-trained graph embeddings with learned word embeddings, we mitigate the inconsistency between predicted relations and given entities. Experimental results on various relation extraction benchmarks in different fields show that knowledge labels improve the F1 score by 1.6% and 1.1% on average over the baseline on standard- and minority-shot benchmarks, respectively.

SUSIE: Pharmaceutical CMC ontology-based information extraction for drug development using machine learning

Automatically extracting information from unstructured text in pharmaceutical documents is important for drug discovery and development. This information can be integrated with structured datasets to ultimately accelerate pharmaceutical product development. To this end, we report an end-to-end information extraction framework based on a custom-built pharmaceutical drug development ontology, a weak supervision framework, contextualization algorithms, and a fine-tuned BioBERT model (adaptation of BERT or Bidirectional Encoder Representations from Transformers for biomedical text). The proposed framework, SUSIE (Schema-based Unsupervised Semantic Information Extraction), was trained on ICH (International Conference on Harmonization) documents to identify important entities and relations from unstructured text and auto-generate knowledge graphs representing crucial information in a structured format. On the entity identification task, the framework achieves a test accuracy and F1-score of 96% and 88%, respectively, on out-of-sample documents. A major contribution of this work is to build an automated, unsupervised information extraction framework around a domain-specific, custom-built pharmaceutical drug development ontology without the need for manual curation of training datasets for specific tasks. The efficacy of the approach was tested on out-of-sample documents including an internal Eli Lilly technical document.

Traditional Chinese medicine for age-related macular degeneration: A clinical evidence map between 2000 and 2022

Abstract Background We systematically searched and summarized clinical studies on traditional Chinese medicine (TCM) for age-related macular degeneration (AMD) with the method of evidence map. We aimed to provide evidence and research direction for the prevention or treatment of AMD with TCM in the future. Materials and methods We searched PubMed, Web of Science, EMBASE, and Chinese databases including CNKI (Chinese National Knowledge Infrastructure), WanFang and VIP, Clinical Trials.gov, and Chinese Clinical Trial Registry Center studies on the prevention or treatment of TCM for AMD from January 2000 to November 2022. We showed distribution features of these studies in the form of text description, graphs, and tables. Results The final 258 articles included 245 clinical studies, 9 systematic reviews or meta-analyses, and 4 guidelines, expert consensus, or clinical pathway. After analyzing the articles with a focus on the stage of AMD, study type, sample size, intervention measures, outcome indicators, and the quality of evidence, we found that the studies were increasing, most of which were wet AMD, the main study type was randomized controlled trials, and the sample size was mostly fewer than 100 cases. Different interventions of TCM, such as Chinese medicine, Chinese patent medicine, or acupuncture, showed the potential benefits of improving vision, reducing fundus lesions, and alleviating some uncomfortable symptoms. However, the sample size of these studies was small. The outcome indicators were not unified, and the stage of AMD was unclear. The quality of articles was lower, and there were few studies on early AMD. The advantages of TCM for the prevention or treatment of AMD remained unclear. Conclusions Early intervention will be the focus and direction of AMD in the prevention or treatment of TCM, and more standardized clinical studies of TCM for early AMD need to be designed and carried out in the future.

Document-level relation extraction based on sememe knowledge-enhanced abstract meaning representation and reasoning

Document-level relation extraction is a challenging task in information extraction, as it involves identifying semantic relations between entities that are dispersed throughout a document. Existing graph-based approaches often rely on simplistic methods to construct text graphs, which do not provide enough lexical and semantic information to accurately predict the relations between entity pairs. In this paper, we introduce a document-level relation extraction method called SKAMRR (Sememe Knowledge-enhanced Meaning Representation and Reasoning). First, we generate document-level abstract meaning representation graphs using rules and acquire entity nodes’ features through sufficient information propagation. Next, we construct inference graphs for entity pairs and utilize graph neural networks to obtain their representations for relation classification. Additionally, we propose the global adaptive loss to address the issue of long-tailed data. We conduct extensive experiments on four datasets DocRE, CDR, GDA, and HacRED. Our model achieves competitive results and its performance outperforms previous state-of-the-art methods on four datasets.

Enhancing Text Classification by Graph Neural Networks With Multi-Granular Topic-Aware Graph

Text classification based on graph neural networks (GNNs) has been widely studied by virtue of its potential to capture complex and across-granularity relations among texts of different types from learning on a text graph. Existing methods typically construct text graphs based on words-documents to capture relevant intra-class document representations among the same documents via words-words and words-documents propagation. However, a natural problem is that polysemy words in documents may become an information medium between documents of different categories, promoting heterophily information propagation. The performance of text classification will be somewhat constrained by this issue. This paper proposes a novel text classification method based on GNN from multi-granular topic-aware perspective, referred to as Text-MGNN. Specifically, topic nodes are introduced to build a triple node set of “word, document, topic,” and multi-granularity relations are modeled on a text graph for this triple node set. The introduction of topic nodes has three significant advantages. The first is to strengthen the propagation of topics, words, and documents. The second is to enhance class-aware representation learning. The final is to mitigate the effect of heterophily information caused by polysemy words. Extensive experiments are conducted on three real-world datasets. Results validate that our proposed method outperforms 11 baselines methods.

The Study on the Text Classification Based on Graph Convolutional Network and BiLSTM

Graph Convolutional Neural Network (GCN) is widely used in text classification tasks. Furthermore, it has been effectively used to accomplish tasks that are thought to have a rich relational structure. However, due to the sparse adjacency matrix constructed by GCN, GCN cannot make full use of context-dependent information in text classification, and it is not good at capturing local information. The Bidirectional Encoder Representation from Transformers (BERT) has the ability to capture contextual information in sentences or documents, but it is limited in capturing global (the corpus) information about vocabulary in a language, which is the advantage of GCN. Therefore, this paper proposes an improved model to solve the above problems. The original GCN uses word co-occurrence relationships to build text graphs. Word connections are not abundant enough and cannot capture context dependencies well, so we introduce a semantic dictionary and dependencies. While the model enhances the ability to capture contextual dependencies, it lacks the ability to capture sequences. Therefore, we introduced BERT and Bi-directional Long Short-Term Memory (BiLSTM) Network to perform deeper learning on the features of text, thereby improving the classification effect of the model. The experimental results show that our model is more effective than previous research reports on four text classification datasets.

Contrastive Graph Convolutional Networks with adaptive augmentation for text classification

Text classification is an important research topic in natural language processing (NLP), and Graph Neural Networks (GNNs) have recently been applied in this task. However, in existing graph-based models, text graphs constructed by rules are not real graph data and introduce massive noise. More importantly, for fixed corpus-level graph structure, these models cannot sufficiently exploit the labeled and unlabeled information of nodes. Meanwhile, contrastive learning has been developed as an effective method in graph domain to fully utilize the information of nodes. Therefore, we propose a new graph-based model for text classification named CGA2TC, which introduces contrastive learning with an adaptive augmentation strategy into obtaining more robust node representation. First, we explore word co-occurrence and document word relationships to construct a text graph. Then, we design an adaptive augmentation strategy for the text graph with noise to generate two contrastive views that effectively solve the noise problem and preserve essential structure. Specifically, we design noise-based and centrality-based augmentation strategies on the topological structure of text graph to disturb the unimportant connections and thus highlight the relatively important edges. As for the labeled nodes, we take the nodes with same label as multiple positive samples and assign them to anchor node, while we employ consistency training on unlabeled nodes to constrain model predictions. Finally, to reduce the resource consumption of contrastive learning, we adopt a random sample method to select some nodes to calculate contrastive loss. The experimental results on several benchmark datasets can demonstrate the effectiveness of CGA2TC on the text classification task.

Conducting a representative national randomized control trial of tailored clinical decision support for nurses remotely: Methods and implications

Clinical Decision Support (CDS) systems, patient specific evidence delivered to clinicians via the electronic health record (EHR) at the right time and in the right format, has the potential to improve patient outcomes. Unfortunately, outcomes of CDS research are mixed. A potential cause lies in its testing. Many CDS are implemented in practice without sufficient testing, potentially leading to patient harm. When testing is conducted, most research has focused on “what” evidence to provide with little attention to the impact of the CDS display format (e.g., textual, graphical) on the user.In an adequately powered randomized control trial with 220 hospital based registered nurses, we will compare 4 randomly assigned CDS format groups (text, text table, text graphs, tailored to subject's graph literacy score) for effects on decision time and simulated patient outcomes. We recruit using state based professional registries, which allows access to participants from multiple institutions across the nation. We use online survey software (REDCap) for efficient study workflow including screening, informed consent documentation, pre-experiment demographic data collection including a graph literacy questionnaire used in randomization. The CDS prototype is accessed via a web app and the simulation-based experiment is conducted remotely at a subject's local computer using video-conferencing software. Also included are 6 post intervention surveys to assess cognitive workload, usability, numeracy, format preference, CDS utilization rationale, and CDS interpretation. Our methods are replicable and scalable for testing of health information technologies and have the potential to improve the safety and effectiveness of these technologies across disciplines.

Word‐Problem Performance Differences by Schema: A Comparison of Students with and without Mathematics Difficulty

AbstractWord‐problem features such as text complexity, charts and graphs, position of the unknown, calculation complexity, irrelevant information, and schemas impact word‐problem performance. We compared the word‐problem performance of typically achieving (TA) students and students with mathematics difficulty (MD). First, we measured the word‐problem performance of all students for schemas and position of the unknown, followed by the performance of students with MD for schemas, position of the unknown, irrelevant information, and charts or graphs. Across schemas, while TA students outperformed students with MD, all students typically scored higher on Change and Difference problems than on Total problems. For position of the unknown, students often scored highest on problems with the final position unknown. Students with MD also demonstrated higher scores on problems with irrelevant information than charts and graphs. Although patterns emerged, not all problems followed the same trends, suggesting the need for further research to investigate the impact of word‐problem features on word‐problem accuracy.

STonKGs: a sophisticated transformer trained on biomedical text and knowledge graphs.

MotivationThe majority of biomedical knowledge is stored in structured databases or as unstructured text in scientific publications. This vast amount of information has led to numerous machine learning-based biological applications using either text through natural language processing (NLP) or structured data through knowledge graph embedding models. However, representations based on a single modality are inherently limited.ResultsTo generate better representations of biological knowledge, we propose STonKGs, a Sophisticated Transformer trained on biomedical text and Knowledge Graphs (KGs). This multimodal Transformer uses combined input sequences of structured information from KGs and unstructured text data from biomedical literature to learn joint representations in a shared embedding space. First, we pre-trained STonKGs on a knowledge base assembled by the Integrated Network and Dynamical Reasoning Assembler consisting of millions of text-triple pairs extracted from biomedical literature by multiple NLP systems. Then, we benchmarked STonKGs against three baseline models trained on either one of the modalities (i.e. text or KG) across eight different classification tasks, each corresponding to a different biological application. Our results demonstrate that STonKGs outperforms both baselines, especially on the more challenging tasks with respect to the number of classes, improving upon the F1-score of the best baseline by up to 0.084 (i.e. from 0.881 to 0.965). Finally, our pre-trained model as well as the model architecture can be adapted to various other transfer learning applications.Availability and implementationWe make the source code and the Python package of STonKGs available at GitHub (https://github.com/stonkgs/stonkgs) and PyPI (https://pypi.org/project/stonkgs/). The pre-trained STonKGs models and the task-specific classification models are respectively available at https://huggingface.co/stonkgs/stonkgs-150k and https://zenodo.org/communities/stonkgs.Supplementary information Supplementary data are available at Bioinformatics online.

Graph Fusion Network for Text Classification

Text classification is an important and classical problem in natural language processing. Recently, Graph Neural Networks (GNNs) have been widely applied in text classification and achieved outstanding performance. Despite the success of GNNs on text classification, existing methods are still limited in two main aspects. On the one hand, transductive methods cannot easily adapt to new documents. Since transductive methods incorporate all documents into their text graph, they need to reconstruct the whole graph and retrain their system from scratch when new documents come. However, this is not applicable to real-world situations. On the other hand, many state-of-the-art algorithms ignore the quality of text graphs, which may lead to sub-optimal performance. To address these problems, we propose a Graph Fusion Network (GFN), which can overcome these limitations and boost text classification performance. In detail, in the graph construction stage, we build homogeneous text graphs with word nodes, which makes the learning system capable of making inference on new documents without rebuilding the whole text graph. Then, we propose to transform external knowledge into structural information and integrate different views of text graphs to capture more structural information. In the graph reasoning stage, we divide the process into three steps: graph learning, graph convolution, and graph fusion. In the graph learning step, we adopt a graph learning layer to further adapt text graphs. In the graph fusion step, we design a multi-head fusion module to integrate different opinions. Experimental results on five benchmarks demonstrate the superiority of our proposed method.

MecCog: a knowledge representation framework for genetic disease mechanism.

Experimental findings on genetic disease mechanisms are scattered throughout the literature and represented in many ways, including unstructured text, cartoons, pathway diagrams and network graphs. Integration and structuring of such mechanistic information greatly enhances its utility. MecCog is a graphical framework for building integrated representations (mechanism schemas) of mechanisms by which a genetic variant causes a disease phenotype. A MecCog mechanism schema displays the propagation of system perturbations across stages of biological organization, using graphical notations to symbolize perturbed entities and activities, hyperlinked evidence tagging, a mechanism ontology and depiction of knowledge gaps, ambiguities and uncertainties. The web platform enables a user to construct, store, publish, browse, query and comment on schemas. MecCog facilitates the identification of potential biomarkers, therapeutic intervention sites and critical future experiments. The MecCog framework is freely available at http://www.meccog.org. Supplementary data are available at Bioinformatics online.

An Investigation of Physics Teachers’ Multiple Representation Ability on Newton’s Law Concept

The ability of the teacher to represent a concept using various types of representation is a critical competency. Teachers who have limited multiple representation abilities will not be able to provide a holistic and appropriate learning experience because students have different abilities in constructing their knowledge. This study aims to determine the multiple representation ability of Physics teacher by using instruments to test multiple representations on the topic of Newton’s law consisting of text representations, images, pictorial diagrams, tables, graphs, and mathematical equations. The subjects in this study were 30 high school Physics teachers in North Maluku who were determined based on the purposive sampling technique. Data on multiple teacher representation abilities were analyzed using descriptive statistics. The results of the data analysis showed that the multi-representation ability of high school physics teachers in Newton’s law concept was in the medium category (55.7%) with the highest achievement occurring in the type of mathematical representation and the lowest in the pictorial diagram representation types for both female and male teachers. Besides, based on data implemented using SPSS 2.0, it can be concluded that there are no significant differences with the ability of teacher’s multi-representation based on gender and teacher’s teaching experience.

Traditional Chinese Medicine for Essential Hypertension: A Clinical Evidence Map

We systematically retrieved and summarised clinical studies on traditional Chinese medicine (TCM) for the prevention and treatment of essential hypertension (EH) using the evidence map. We aimed to explore the evidence distribution, identify gaps in evidence, and inform on future research priorities. Clinical studies, systematic reviews, guidelines, and pathway studies related to TCM for the prevention and treatment of EH, published between January 2000 and December 2019, were included from databases CNKI, WanFang Data, VIP, PubMed, Embase, and Web of Science. The distribution of evidence was analysed using text descriptions, tables, and graphs. A total of 9,403 articles were included, including 5,920 randomised controlled studies (RCTs), 16 guidelines, expert consensus and path studies, and 139 systematic reviews (SRs). The articles publishing trend increased over time. This study showed that the intervention time of TCM was concentrated at 4–8 weeks, mainly through Chinese herbal medicine (CHM) for the prevention and treatment of elderly hypertension and the complications. A Measurement Tool to Assess Systematic Reviews (AMSTAR) scores of the included reviews ranged from 2 to 10. Most of the SRs had a potentially positive effect (n = 120), mainly in 5–8 score. Primary studies and SRs show potential benefits of TCM in lowering blood pressure, lowering the TCM syndrome and symptom differentiation scores (TCM-SSD scores), improving the total effective rate, and reducing the adverse events. The adjunctive effect of TCM on improving the total effective rate, lowering the blood pressure, lowering the TCM-SSD scores, and lowering the adverse effects was only supported by low-quality evidence in this research. The evidence map was used to show the overall research on TCM for the treatment of EH; however, due to the existing problems of the primary studies, the current research conclusion needs further research with higher quality and standardisation.

Растительность: отображение в новом «Экологическом атласе России»

In 2017 the fundamental scientific-reference multidisciplinary Ecological Atlas of Russia was published (Ecological …, 2017; Kasimov et al., 2018). The Atlas reflects the ecological situation at the beginning of the 21st century. The Geography Department of Lomonosov Moscow State University with the participation of more than 30 leading departmental and scientific organizations contributed to the Atlas. The Atlas represents a wide range of ecological-geographical spatio-temporal characteristics of the territory of Russia and its regions. The six structural sections of the Atlas contain more than 30 maps showing vegetation in different aspects: Introduction; Natural conditions for the formation of an ecological situation; The impact of economic activity on the environment; Natural and technological hazards; Modern ecological situation; Environmental monitoring and nature conservation. The scale of the base maps of Russia is 1 : 20 000 000, others — 1 : 30 000 000 and smaller. Maps are accompanied by text descriptions, graphs and slides. More than 20 % of the Atlas volume is given to satellite imagery — an effective, in some cases unique, means of visualizing environmental information. The description of the maps is given in the following sequence: inventory maps — estimation maps. The Introduction “Russia on the Ecological Map of the World” analyzes the ecological role of Russia on a planetary scale and assesses the contribution to the observed degradation of the planet’s environment. The text reveals the role of vegetation in the biosphere and its environmental functions. In the section “Natural Conditions for the Formation of an Ecological Situation” there is a photomap “Vegetation Cover” created using MODIS images. The 18 divisions of vegetation are grouped in the legend into five large typological complexes — Forests, Grass and shrub vegetation, Tundra, Wetland complexes, Other vegetation. Mires are represented by three maps in 1 : 30 000 000 scale: “Mires and wetlands” (Fig. 1), “Types of mires”, “Afforestation of mires”. The key topic ‒ “Ecological functions of the vegetation cover” — has been made as a separate map (Volkova, Fedorova, 1995). Large proportion of the section is devoted to the productivity of the vegetation cover (3 maps), the most important indicator controlling the stability of geosystems (Fig. 2). In the section “Impact of economic activity on the environment”, vegetation is displayed through the main object of economic activity — forests and factors that determine the current state of forests: deforestation, derivative forests, forest burnability, and frequency of forest fires. The cumulative effect of their impact is presented on the map “Forest disturbance” (Fig. 3). The consequences of adverse effects on biota are presented on the integrated map “Degradation of the plant and animal world” at a scale of 1 : 20 000 000. The maps of poisonous plants and plants-allergens in 1 : 30 000 000 scale (Dikareva et al., 2017) were made for the first time; they are placed in the section “Natural and technological hazards” (Fig. 4). The map “Ecological state of natural fodder lands” (1 : 20 000 000 s.) is included in the group of maps characterizing the ecological state of individual natural components (surface and underground waters, soils, lands, etc.). The final section of the Atlas “Environmental monitoring and nature conservation” contains the maps “Nature Protection”, “Specially Protected Natural Territories”, “Especially Valuable Wetlands” and maps of the Red Book species of plants. The section concludes with the topic “Environmental Benefits of the Russian Federation and Their Capitalization. Russia is in the market of ecosystem services”. It complements the Introduction chapter, focusing on the huge role of the territory of Russia as a natural regulator of the global environment and the need to capitalize its environmental benefits.

HRGRN: A Graph Search-Empowered Integrative Database of Arabidopsis Signaling Transduction, Metabolism and Gene Regulation Networks.

The biological networks controlling plant signal transduction, metabolism and gene regulation are composed of not only tens of thousands of genes, compounds, proteins and RNAs but also the complicated interactions and co-ordination among them. These networks play critical roles in many fundamental mechanisms, such as plant growth, development and environmental response. Although much is known about these complex interactions, the knowledge and data are currently scattered throughout the published literature, publicly available high-throughput data sets and third-party databases. Many ‘unknown’ yet important interactions among genes need to be mined and established through extensive computational analysis. However, exploring these complex biological interactions at the network level from existing heterogeneous resources remains challenging and time-consuming for biologists. Here, we introduce HRGRN, a graph search-empowered integrative database of Arabidopsis signal transduction, metabolism and gene regulatory networks. HRGRN utilizes Neo4j, which is a highly scalable graph database management system, to host large-scale biological interactions among genes, proteins, compounds and small RNAs that were either validated experimentally or predicted computationally. The associated biological pathway information was also specially marked for the interactions that are involved in the pathway to facilitate the investigation of cross-talk between pathways. Furthermore, HRGRN integrates a series of graph path search algorithms to discover novel relationships among genes, compounds, RNAs and even pathways from heterogeneous biological interaction data that could be missed by traditional SQL database search methods. Users can also build subnetworks based on known interactions. The outcomes are visualized with rich text, figures and interactive network graphs on web pages. The HRGRN database is freely available at http://plantgrn.noble.org/hrgrn/.