Graph Theoretic Clustering Algorithm Research Articles

Identifying correlation clusters in many-body localized systems

We introduce techniques for analysing the structure of quantum states of many-body localized (MBL) spin chains by identifying correlation clusters from pairwise correlations. These techniques proceed by interpreting pairwise correlations in the state as a weighted graph, which we analyse using an established graph theoretic clustering algorithm. We validate our approach by studying the eigenstates of a disordered XXZ spin chain across the MBL to ergodic transition, as well as the non-equilibrium dyanmics in the MBL phase following a global quantum quench. We successfully reproduce theoretical predictions about the MBL transition obtained from renormalization group schemes. Furthermore, we identify a clear signature of many-body dynamics analogous to the logarithmic growth of entanglement. The techniques that we introduce are computationally inexpensive and in combination with matrix product state methods allow for the study of large scale localized systems. Moreover, the correlation functions we use are directly accessible in a range of experimental settings including cold atoms.

KNN-SC: Novel Spectral Clustering Algorithm Using k-Nearest Neighbors

Spectral clustering is a well-known graph-theoretic clustering algorithm. Although spectral clustering has several desirable advantages (such as the capability of discovering non-convex clusters and applicability to any data type), it often leads to incorrect clustering results because of high sensitivity to noise points. In this study, we propose a robust spectral clustering algorithm known as KNN-SC that can discover exact clusters by decreasing the influence of noise points. To achieve this goal, we present a novel approach that filters out potential noise points by estimating the density difference between data points using $k$ -nearest neighbors. In addition, we introduce a novel method for generating a similarity graph in which various densities of data points are effectively represented by expanding the nearest neighbor graph. Experimental results on synthetic and real-world datasets demonstrate that KNN-SC achieves significant performance improvement over many state-of-the-art spectral clustering algorithms.

Nature-effect relationship research of Curcumae Longae Rhizoma,Curcumae Radix,and Curcumae Rhizoma based on nature combination

Curcumae Longae Rhizoma,Curcumae Radix and Curcumae Rhizoma are different medicinal parts of the same plant.Nevertheless,they are different in medicinal effects due to the different Chinese herbal nature. In this study,traditional Chinese medicines database( TCMD2009),traditional Chinese medicine system( TCMSP),and Ch EMBL database were retrieved to screen the active components and targets,and construct the target PPI network. By a graph theoretic clustering algorithm identifying protein complex algorithm( IPCA),the protein modules were identified and analyzed by gene ontology( GO) enrichment. A comparative study of Curcumae Longae Rhizoma and Curcumae Radix illustrate that Curcumae Longae Rhizoma regulates blood coagulation through P2 RY12,GNG2 and other genes to exert the analgesic effect. Curcumae Radix regulates lipid metabolism,plasma lipoprotein particle levels,platelet activation,response to oxidative stress,apoptotic process through LDLR,APOB,PRKCA,SOD1,TP53 and other genes to perform a function in clearing the heart and cooling the blood. A comparative study of Curcumae Radix and Curcumae Rhizoma demonstrate that Curcumae Rhizoma on regulates the nervous system by GRIA2,GRIA4 and other genes to exert blood-breaking effect; Curcumae Radix regulates lipid metabolism,plasma lipoprotein particle levels,platelet activation,response to oxidative stress,apoptotic process by genes such as CALM1,LPL,APOB,SOD1 and TP53 to play the role of clearing heart and cooling blood. Cluster analysis of the protein interaction network of the nature combination is helpful to explain the intrinsic link between the nature combination and efficacy.

Modified Graph-theoretic Clustering Algorithm for Mining International Linkages of Philippine Higher Education Institutions

Graph-theoretic clustering either uses limited neighborhood or construction of a minimum spanning tree to aid the clustering process. The latter is challenged by the need to identify and consequently eliminate inconsistent edges to achieve final clusters, detect outliers and partition substantially. This work focused on mining the data of the International Linkages of Philippine Higher Education Institutions by employing a modified graph-theoretic clustering algorithm with which the Prim’s Minimum Spanning Tree algorithm was used to construct a minimum spanning tree for the internationalization dataset infusing the properties of a small world network. Such properties are invoked by the computation of local clustering coefficient for the data elements in the limited neighborhood of data points established using the von Neumann Neighborhood. The overall result of the cluster validation using the Silhouette Index with a score of .69 indicates that there is an acceptable structure found in the clustering result – hence, a potential of the modified MST-based clustering algorithm. The Silhouette per cluster with .75 being the least score means that each cluster derived for r=5 by the von Neumann Neighborhood has a strong clustering structure.

A novel clustering method to identify cell types from single cell transcriptional profiles

With the advancement of the high throughput single cell techniques, transcriptomics data generation at the single cell level becomes very easy. Analysis of this single cell expression values can reveal lots of unprecedented information about complex cellular heterogeneity and tissue composition. To date, different statistical methods are applied to analyze expression data at the cellular level, but there is still pretty much scope for the development of new bioinformatics tools. In this article, a graph theoretic clustering algorithm is proposed to identify cellular states from single cell gene expression data. The proposed algorithm first generates a shared nearest neighbor graph from the single cell RNA-seq dataset and then applies minimum spanning tree based clustering method to cluster the graph nodes. To compare our proposed algorithm’s performance with other unsupervised clustering methods, we used three real scRNA-seq datasets (human cancer cells, human embryonic cells and mouse embryonic cells). From the comparison result, it is evident that the proposed algorithm outperforms other standard single cell analysis methods.

Pharmacological mechanism analysis of oligopeptide from Pinctada fucata based on in silico proteolysis and protein interaction network

Pinctada fucata oligopeptide is one of key pharmaceutical effective constituents of P. fucata. It is significant to analyze its pharmacological effect and mechanism. This study aims to discover the potential oligopeptides from P. fucata and analyze the mechanism of P. fucata oligopeptide based on in silico technologies and protein interaction network(PIN). First, main protein sequences of P. fucata were collected, and oligopeptides were obtained using in silico gastrointestinal tract proteolysis. Then, key potential targets of P. fucata oligopeptides were obtained through pharmacophore screening. The protein-protein interaction(PPI) of targets was achieved and implemented to construct PIN and analyze the mechanism of P. fucata oligopeptides. P. fucata oligopeptide database was constructed based on in silico technologies, including 458 oligopeptides. Twelve modules were identified from PIN by a graph theoretic clustering algorithm Molecular Complex Detection(MCODE) and analyzed by Gene ontology(GO) enrichment. The results indicated that P. fucata oligopeptides have an effect in treating neurological diseases, such as Alzheimer's disease. In silico proteolysis could be used to analyze the protein sequences of traditional Chinese medicine(TCM). According to the combination of in silico proteolysis and PIN, the biological activity of oligopeptides could be interpreted rapidly based on the known TCM protein sequence. The study provides the methodology basis for rapidly and efficiently implementing the mechanism analysis of TCM oligopeptides.

Anti-inflammatory mechanism analysis of antirheumatic Chinese medicinal herb Aconiti Radix based on protein interaction network

Inflammation is one of the important risk factors of rheumatic diseases. Aconiti Radix is widely used for the treatment of rheumatism, which has significant anti-inflammatory effects. However, its anti-inflammatory mechanism on molecular level is still not clear. The purpose of this study is to illuminate the anti-inflammatory mechanism of Aconiti Radix based on the protein interaction network (PIN) analysis on molecular network level. The main anti-inflammatory components (aconitine, hypaconitine and mesaconitine) were chosen in this study to obtain the targets of the components and protein-protein information though databases retrieval and construct the PIN of Aconiti Radix. By a graph theoretic clustering algorithm molecular complex detection(MCODE), 13 modules were identified and analyzed by gene ontology(GO) enrichment. The results showed that the anti-inflammatory mechanism of Aconiti Radix was mainly associated with prostanoid metabolic process and leukocyte chemotaxis mediated by chemokines. In this study, the anti-inflammatory mechanism of Aconiti Radix was elucidated systematically from molecular network level, which provided the scientific basis for the treatment of rheumatic diseases.

Anti-inflammatory mechanism research of flavonoid compounds in Dalbergiae Odoriferae Lignum by module-based network analysis

Dalbergiae Odoriferae Lignum as a traditional Chinese medicine (TCM) has been widely used for promoting blood circulation and removing blood stasis. Flavonoid compounds are main chemical constituents of Dalbergiae Odoriferae Lignum, which exert anti-inflammatory property. However, the underlying anti-inflammatory mechanisms of flavonoid compounds are incompletely understood. It has been reported that isoliquiritigenin, liquiritigenin, naringenin and butein possess anti-inflammatory property. The purpose of this study is to illuminate the anti-inflammatory mechanism of flavonoid compounds based on the protein interaction network (PIN) analysis on molecular network level. 130 targets of the main medicinal ingredients of flavonoid compounds were gained though database retrieval. A protein interaction network of flavonoid compounds was constructed with 589 nodes and 216 interactions. By a graph theoretic clustering algorithm Molecular Complex Detection (MCODE), 26 modules were identified and analyzed by Gene ontology (GO) enrichment. Two modules were associated with anti-inflammatory actions. The most interesting finding of this study was that the anti-inflammatory effect of flavonoid compounds may be partly attributable to inhibite FOS, PTGS2 expression, inhibite of IL-1beta release, and block the MAPK pathway and toll-like receptor pathway.

CLEAR: Class Level Software Refactoring Using Evolutionary Algorithms

AbstractThe original design of a software system is rarely prepared for every new requirement. Software systems should be updated frequently, which is usually accompanied by the decline in software modularity and quality. Although many approaches have been proposed to improve the quality of software, a majority of them are guided by metrics defined on the local properties of software. In this article, we propose to use a global metric borrowed from the network science to detect the moving method refactoring. First, our approach uses a bipartite network to represent classes, features (i.e., methods and fields), and their dependencies. Second, a new metric is introduced to quantify the modularity of a software system as a whole. Finally, a crossover-only evolutionary algorithm that uses the metric as its fitness function is introduced to optimize the class structure of a software system and detect the methods that should be moved. Empirical results on the benchmark Java projects show that our approach can find meaningful methods that should be moved with a high stability. The advantages of our approach are illustrated in comparison with some other approaches, specifically one refactoring approach, namely search-based refactoring approach (SBRA), and two community detection algorithms, namely a graph theoretic clustering algorithm (MCODE) and a fast algorithm for community detection (FG). Our approach provides a new way to do refactoring from the perspective of software structure.

Protein interaction network analysis of Panax notoginseng saponins

Panax notoginseng (PN) is one of the commonly used clinical medicines for cardiovascular diseases and possesses a variety of pharmacological effects. P. notoginseng saponins (PNS) are the most important bioactive components in PN. The purpose of this study was to explain the mechanism of PNS on molecular network level. 18 targets of the main medicinal ingredients of PNS were gained by virtual screening based on pharmacophores and data mining. A protein interaction network of PNS was constructed with 189 nodes and 721 interactions. By a graph theoretic clustering algorithm Molecular Complex Detection (MCODE), 14 modules were detected. Gene ontology (GO) enrichment analysis of the modules demonstrated that the roles of PNS played in cardiovascular disease related to multiple biological processes, which could represent the characteristics of traditional Chinese medicine (TCM) as a whole to regulate the disease. The results showed that the blood circulation and hemostasis efficacy of PN related with the biological processes such as positive regulation of cAMP metabolic and biosynthetic process, platelet activation and regulation of blood vessel size, regulation of T cell proliferation and differentiation and so on. Therefore, the module-based network analysis will be an effective method for better understanding TCM.

Effectiveness of Partition and Graph Theoretic Clustering Algorithms for Multiple Source Partial Discharge Pattern Classification Using Probabilistic Neural Network and Its Adaptive Version: A Critique Based on Experimental Studies

Partial discharge (PD) is a major cause of failure of power apparatus and hence its measurement and analysis have emerged as a vital field in assessing the condition of the insulation system. Several efforts have been undertaken by researchers to classify PD pulses utilizing artificial intelligence techniques. Recently, the focus has shifted to the identification of multiple sources of PD since it is often encountered in real‐time measurements. Studies have indicated that classification of multi‐source PD becomes difficult with the degree of overlap and that several techniques such as mixed Weibull functions, neural networks, and wavelet transformation have been attempted with limited success. Since digital PD acquisition systems record data for a substantial period, the database becomes large, posing considerable difficulties during classification. This research work aims firstly at analyzing aspects concerning classification capability during the discrimination of multisource PD patterns. Secondly, it attempts at extending the previous work of the authors in utilizing the novel approach of probabilistic neural network versions for classifying moderate sets of PD sources to that of large sets. The third focus is on comparing the ability of partition‐based algorithms, namely, the labelled (learning vector quantization) and unlabelled (K‐means) versions, with that of a novel hypergraph‐based clustering method in providing parsimonious sets of centers during classification.

Protein complex prediction based on mutually exclusive interactions in protein interaction network.

The increasing amount of available Protein-Protein Interaction (PPI) data enables scalable methods for the protein complex prediction. A protein complex is a group of two or more proteins formed by interactions that are stable over time, and it generally corresponds to a dense sub-graph in PPI Network (PPIN). However, dense sub-graphs correspond not only to stable protein complexes but also to sets of proteins including dynamic interactions. As a result, conventional simple PPIN based graph-theoretic clustering methods have high false positive rates in protein complex prediction. In this paper, we propose an approach to predict protein complexes based on the integration of PPI data and mutually exclusive interaction information drawn from structural interface data of protein domains. The extraction of Simultaneous Protein Interaction Cluster (SPIC) is the essence of our approach, which excludes interaction conflicts in network clusters by achieving mutually exclusion among interactions. The concept of SPIC was applied to conventional graph-theoretic clustering algorithms, MCODE and LCMA, to evaluate the density of clusters for protein complex prediction. The comparison with original graph-theoretic clustering algorithms verified the effectiveness of our approach; SPIC based methods refined false positives of original methods to be true positive complexes, without any loss of true positive predictions yielded by original methods.

IDENTIFICATION OF FUNCTIONAL MODULES IN A PPI NETWORK BY BOUNDED DIAMETER CLUSTERING

Dense subgraphs of Protein-Protein Interaction (PPI) graphs are assumed to be potential functional modules and play an important role in inferring the functional behavior of proteins. Increasing amount of available PPI data implies a fast, accurate approach of biological complex identification. Therefore, there are different models and algorithms in identifying functional modules. This paper describes a new graph theoretic clustering algorithm that detects densely connected regions in a large PPI graph. The method is based on finding bounded diameter subgraphs around a seed node. The algorithm has the advantage of being very simple and efficient when compared with other graph clustering methods. This algorithm is tested on the yeast PPI graph and the results are compared with MCL, Core-Attachment, and MCODE algorithms.

Graph Matching-Based Distributed Clustering and Backbone Formation Algorithms for Sensor Networks

Clustering is a widely used technique to manage the essential operations such as routing and data aggregation in wireless sensor networks (WSNs). We propose two new graph-theoretic distributed clustering algorithms for WSNs that use a weighted matching method for selecting strong links. To the best of our knowledge, our algorithms are the first attempts that use graph matching for clustering. The first algorithm is divided into rounds; extended weighted matching operation is executed by nodes in each round; thus the clusters are constructed synchronously. The second algorithm is the enhanced version of the first algorithm, which provides not only clustering but also backbone formation in an energy-efficient and asynchronous manner. We show the operation of the algorithms, analyze them, provide the simulation results in an ns2 environment. We compare our proposed algorithms with the other graph-theoretic clustering algorithms and show that our algorithms select strong communication links and create a controllable number of balanced clusters while providing low-energy consumptions. We also discuss possible applications that may use the structure provided by these algorithms and the extensions to the algorithms.

An approach and an Eclipse-based environment for enhancing the navigation structure of Web sites

This paper presents an approach based on information retrieval and clustering techniques for automatically enhancing the navigation structure of a Web site for improving navigability. The approach increments the set of navigation links provided in each page of the site with a semantic navigation map, i.e., a set of links enabling navigating from a given page to other pages of the site showing similar or related content. The approach uses Latent Semantic Indexing to compute a dissimilarity measure between the pages of the site and a graph-theoretic clustering algorithm to group pages showing similar or related content according to the calculated dissimilarity measure. AJAX code is finally used to extend each Web page with an associated semantic navigation map. The paper also presents a prototype of a tool developed to support the approach and the results from a case study conducted to assess the validity and feasibility of the proposal.

Unsupervised Active Learning Based on Hierarchical Graph-Theoretic Clustering

Most existing active learning approaches are supervised. Supervised active learning has the following problems: inefficiency in dealing with the semantic gap between the distribution of samples in the feature space and their labels, lack of ability in selecting new samples that belong to new categories that have not yet appeared in the training samples, and lack of adaptability to changes in the semantic interpretation of sample categories. To tackle these problems, we propose an unsupervised active learning framework based on hierarchical graph-theoretic clustering. In the framework, two promising graph-theoretic clustering algorithms, namely, dominant-set clustering and spectral clustering, are combined in a hierarchical fashion. Our framework has some advantages, such as ease of implementation, flexibility in architecture, and adaptability to changes in the labeling. Evaluations on data sets for network intrusion detection, image classification, and video classification have demonstrated that our active learning framework can effectively reduce the workload of manual classification while maintaining a high accuracy of automatic classification. It is shown that, overall, our framework outperforms the support-vector-machine-based supervised active learning, particularly in terms of dealing much more efficiently with new samples whose categories have not yet appeared in the training samples.

BAG: a graph theoretic sequence clustering algorithm

In this paper, we first discuss issues in clustering biological sequences with graph properties, which inspired the design of our sequence clustering algorithm BAG. BAG recursively utilises several graph properties: biconnectedness, articulation points, pquasi-completeness, and domain knowledge specific to biological sequence clustering. To reduce the fragmentation issue, we have developed a new metric called cluster utility to guide cluster splitting. Clusters are then merged back with less stringent constraints. Experiments with the entire COG database and other sequence databases show that BAG can cluster a large number of sequences accurately while keeping the number of fragmented clusters significantly low.

CLUE: cluster-based retrieval of images by unsupervised learning

In a typical content-based image retrieval (CBIR) system, target images (images in the database) are sorted by feature similarities with respect to the query. Similarities among target images are usually ignored. This paper introduces a new technique, cluster-based retrieval of images by unsupervised learning (CLUE), for improving user interaction with image retrieval systems by fully exploiting the similarity information. CLUE retrieves image clusters by applying a graph-theoretic clustering algorithm to a collection of images in the vicinity of the query. Clustering in CLUE is dynamic. In particular, clusters formed depend on which images are retrieved in response to the query. CLUE can be combined with any real-valued symmetric similarity measure (metric or nonmetric). Thus, it may be embedded in many current CBIR systems, including relevance feedback systems. The performance of an experimental image retrieval system using CLUE is evaluated on a database of around 60,000 images from COREL. Empirical results demonstrate improved performance compared with a CBIR system using the same image similarity measure. In addition, results on images returned by Google's Image Search reveal the potential of applying CLUE to real-world image data and integrating CLUE as a part of the interface for keyword-based image retrieval systems.

Into the heart of darkness: large-scale clustering of human non-coding DNA.

It is currently believed that the human genome contains about twice as much non-coding functional regions as it does protein-coding genes, yet our understanding of these regions is very limited. We examine the intersection between syntenically conserved sequences in the human, mouse and rat genomes, and sequence similarities within the human genome itself, in search of families of non-protein-coding elements. For this purpose we develop a graph theoretic clustering algorithm, akin to the highly successful methods used in elucidating protein sequence family relationships. The algorithm is applied to a highly filtered set of about 700 000 human-rodent evolutionarily conserved regions, not resembling any known coding sequence, which encompasses 3.7% of the human genome. From these, we obtain roughly 12 000 non-singleton clusters, dense in significant sequence similarities. Further analysis of genomic location, evidence of transcription and RNA secondary structure reveals many clusters to be significantly homogeneous in one or more characteristics. This subset of the highly conserved non-protein-coding elements in the human genome thus contains rich family-like structures, which merit in-depth analysis. Supplementary material to this work is available at http://www.soe.ucsc.edu/~jill/dark.html

A hierarchical multiple classifier learning algorithm

This paper addresses the classification problem for applications with extensive amounts of data and a large number of features. The learning system developed utilizes a hierarchical multiple classifier scheme and is flexible, efficient, highly accurate and of low cost. The system has several novel features: (1) It uses a graph-theoretic clustering algorithm to group the training data into possibly overlapping cluster, each representing a dense region in the data space; (2) component classifiers trained on these dense regions are specialists whose probabilistic outputs are gated inputs to a super-classifier. Only those classifiers whose training clusters are most related to an unknown data instance send their outputs to the super-classifier; and (3) sub-class labelling is used to improve the classification of super-classes. The learning system achieves the goals of reducing the training cost and increasing the prediction accuracy compared to other multiple classifier algorithms. The system was tested on three large sets of data, two from the medical diagnosis domain and one from a forest cover classification problem. The results are superior to those obtained by several other learning algorithms.