SEARCH · PubMed Health

Results for “Clustering Algorithms”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,657 records · Page 92Linked to original sources

PubMed2001

Clusters in alpha/beta barrel proteins: implications for protein structure, function, and folding: a graph theoretical approach.

The alpha/beta barrel fold is adopted by most enzymes performing a variety of catalytic reactions, but with very low sequence similarity. In order to understand the stabilizing interactions important in maintaining the alpha/beta barrel fold, we have identified residue clusters in a dataset of 36 alpha/beta barrel proteins that have less than 10% sequence identity within themselves. A graph theoretical algorithm is used to identify backbone clusters. This approach uses the global information of the nonbonded interaction in the alpha/beta barrel fold for the clustering procedure. The nonbonded interactions are represented mathematically in the form of an adjacency matrix. On diagonalizing the adjacency matrix, clusters and cluster centers are obtained from the highest eigenvalue and its corresponding vector components. Residue clusters are identified in the strand regions forming the beta barrel and are topologically conserved in all 36 proteins studied. The residues forming the cluster in each of the alpha/beta protein are also conserved among the sequences belonging to the same family. The cluster centers are found to occur in the middle of the strands or in the C-terminal of the strands. In most cases, the residues forming the clusters are part of the active site or are located close to the active site. The folding nucleus of the alpha/beta fold is predicted based on hydrophobicity index evaluation of residues and identification of cluster centers. The predicted nucleation sites are found to occur mostly in the middle of the strands. Proteins 2001;43:103-112.

Algorithms↗

PubMed2006

A fuzzy clustering neural network architecture for classification of ECG arrhythmias.

Accurate and computationally efficient means of classifying electrocardiography (ECG) arrhythmias has been the subject of considerable research effort in recent years. This study presents a comparative study of the classification accuracy of ECG signals using a well-known neural network architecture named multi-layered perceptron (MLP) with backpropagation training algorithm, and a new fuzzy clustering NN architecture (FCNN) for early diagnosis. The ECG signals are taken from MIT-BIH ECG database, which are used to classify 10 different arrhythmias for training. These are normal sinus rhythm, sinus bradycardia, ventricular tachycardia, sinus arrhythmia, atrial premature contraction, paced beat, right bundle branch block, left bundle branch block, atrial fibrillation and atrial flutter. For testing, the proposed structures were trained by backpropagation algorithm. Both of them tested using experimental ECG records of 92 patients (40 male and 52 female, average age is 39.75 +/- 19.06). The test results suggest that a new proposed FCNN architecture can generalize better than ordinary MLP architecture and also learn better and faster. The advantage of proposed structure is a result of decreasing the number of segments by grouping similar segments in training data with fuzzy c-means clustering.

Adult↗

PubMed1995

Using the Potts glass for solving the clustering problem.

We present an application of a Potts glass to the clustering problem. Simulated annealing in the mean field approximation is used in order to avoid local minima. The resulting updating equations are completely parallel, and very easy to implement. The model has no free parameters except for the annealing parameters. We show how the model can be implemented for some special clustering problems. The T-->0 limit of the Potts glass is identical to the vector quantization algorithm with certain increments. A comparative study of the Potts glass and vector quantization is also made, and it is shown that for difficult clustering problems, the Potts glass is far better than vector quantization.

Algorithms↗

PubMed2002

Prospective use of molecular typing of Mycobacterium tuberculosis by use of restriction fragment-length polymorphism in a public tuberculosis-control program.

We performed a prospective, community-based evaluation of molecular typing of Mycobacterium tuberculosis isolates as a method for tuberculosis (TB) control. We performed restriction fragment-length polymorphism (RFLP) analysis of the insertion sequences IS6110 and pTBN12 for isolates recovered from 61 of 62 patients with culture-positive TB in St. Louis during 12 months. Twenty-four (39%) of the 61 patients were infected with an isolate with an RFLP pattern that was shared with >/=1 other isolate, and 11 (46%) also had epidemiologic links with patients in their cluster of cases. One case each of laboratory cross-contamination and occupational transmission were discovered. The patients in clusters were more likely to be younger, black, United States-born, to have substance abuse problems, and to live in poorer areas. A predictive algorithm for molecular identification of clusters had a sensitivity and a specificity of 75%. This study allowed the TB-control program in St. Louis to be redirected toward the affected subpopulations.

Adult↗

PubMed1994

Constructing gene models from accurately predicted exons: an application of dynamic programming.

This paper presents a computationally efficient algorithm, the Gene Assembly Program III (GAP III), for constructing gene models from a set of accurately-predicted 'exons'. The input to the algorithm is a set of clusters of exon candidates, generated by a new version of the GRAIL coding region recognition system. The exon candidates of a cluster differ in their presumed edges and occasionally in their reading frames. Each exon candidate has a numerical score representing its 'probability' of being an actual exon. GAP III uses a dynamic programming algorithm to construct a gene model, complete or partial, by optimizing a predefined objective function. The optimal gene models constructed by GAP III correspond very well with the structures of genes which have been determined experimentally and reported in the Genome Sequence Database (GSDB). On a test set of 137 human and mouse DNA sequences consisting of 954 true exons, GAP III constructed 137 gene models using 892 exons, among which 859 (859/954 = 90%) are true exons and 33 (33/892 = 3%) are false positive. Among the 859 true positives, 635 (74%) match the actual exons exactly, and 838 (98%) have at least one edge correct. GAP III is computationally efficient. If we use E and C to represent the total number of exon candidates in all clusters and the number of clusters, respectively, the running time of GAP III is proportional to (E x C).

Algorithms↗

PubMed2005

HierS: hierarchical scaffold clustering using topological chemical graphs.

An exhaustive ring-based algorithm, HierS, has been developed in order to provide an intuitive approach to compound clustering for analyzing high-throughput screening results. The recursive algorithm rapidly identifies all possible ring-delimited substructures within a set of compounds. Molecules are grouped by shared ring substructures (scaffolds) so that common scaffolds obtain higher membership. Once all of the scaffolds for a set of compounds are identified, the hierarchical structural relationships between the scaffold structures are established. The complex network of hierarchical relationships is then utilized to navigate compounds in a structurally directed fashion. When the scaffold hierarchy is traversed, over-represented structural features can be rapidly identified so that excess compounds that contain them can be removed without significantly impacting the structural diversity landscape of the compound set. Furthermore, the removed compounds can provide the opportunity to follow-up on active compounds that had previously been discarded because of practical limitations on follow-up capacity. A Web-based interface has been developed that incorporates this algorithm in order to allow for an interactive analysis. In addition, biological data are coupled to scaffolds by the inclusion of activity histograms, which indicate how the compounds in each scaffold class performed in previous high-throughput screening campaigns.

Algorithms↗

PubMed2003

Design and analysis of a content-based pathology image retrieval system.

A prototype, content-based image retrieval system has been built employing a client/server architecture to access supercomputing power from the physician's desktop. The system retrieves images and their associated annotations from a networked microscopic pathology image database based on content similarity to user supplied query images. Similarity is evaluated based on four image feature types: color histogram, image texture, Fourier coefficients, and wavelet coefficients, using the vector dot product as a distance metric. Current retrieval accuracy varies across pathological categories depending on the number of available training samples and the effectiveness of the feature set. The distance measure of the search algorithm was validated by agglomerative cluster analysis in light of the medical domain knowledge. Results show a correlation between pathological significance and the image document distance value generated by the computer algorithm. This correlation agrees with observed visual similarity. This validation method has an advantage over traditional statistical evaluation methods when sample size is small and where domain knowledge is important. A multi-dimensional scaling analysis shows a low dimensionality nature of the embedded space for the current test set.

Abstracting and Indexing↗

PubMed2005

Sample phenotype clusters in high-density oligonucleotide microarray data sets are revealed using Isomap, a nonlinear algorithm.

BACKGROUND: Life processes are determined by the organism's genetic profile and multiple environmental variables. However the interaction between these factors is inherently non-linear. Microarray data is one representation of the nonlinear interactions among genes and genes and environmental factors. Still most microarray studies use linear methods for the interpretation of nonlinear data. In this study, we apply Isomap, a nonlinear method of dimensionality reduction, to analyze three independent large Affymetrix high-density oligonucleotide microarray data sets. RESULTS: Isomap discovered low-dimensional structures embedded in the Affymetrix microarray data sets. These structures correspond to and help to interpret biological phenomena present in the data. This analysis provides examples of temporal, spatial, and functional processes revealed by the Isomap algorithm. In a spinal cord injury data set, Isomap discovers the three main modalities of the experiment--location and severity of the injury and the time elapsed after the injury. In a multiple tissue data set, Isomap discovers a low-dimensional structure that corresponds to anatomical locations of the source tissues. This model is capable of describing low- and high-resolution differences in the same model, such as kidney-vs.-brain and differences between the nuclei of the amygdala, respectively. In a high-throughput drug screening data set, Isomap discovers the monocytic and granulocytic differentiation of myeloid cells and maps several chemical compounds on the two-dimensional model. CONCLUSION: Visualization of Isomap models provides useful tools for exploratory analysis of microarray data sets. In most instances, Isomap models explain more of the variance present in the microarray data than PCA or MDS. Finally, Isomap is a promising new algorithm for class discovery and class prediction in high-density oligonucleotide data sets.

Algorithms↗

PubMed1997

Drug-motif-based diverse monomer selection: method and application in combinatorial chemistry.

This article describes a strategy to explore monomer diversity while incorporating drug motif knowledge into the design and selection of monomers for combinatorial chemistry. The process involves collecting from available electronic databases all those molecules that potentially could be monomers. In this manner we have assembled five DAYLIGHT databases, each containing one of the common functional groups: carboxylic acids, aldehydes, nitriles, primary amines, and secondary amines. The molecules in the databases are then subjected to fingerprint and cluster analysis using the Jarvis-Patrick algorithm and profiles of the compounds are calculated relating to molecular weight, H-bond counts, and rotatable bond flexibility. The cluster information and profiles of the molecules are stored back into the databases for similarity and diversity searches, and for profile prescreening of monomers. To apply drug motif knowledge to a selection an application to an aldehyde set is discussed, in which representatives of each cluster in the aldehyde database are compared with drug molecules in the Standard Derwent File (SDF) in one of three ways to select drug motif-based monomers for purchase or synthesis.

Aldehydes↗

PubMed2006

Benchmarking a memetic algorithm for ordering microarray data.

This work introduces a new algorithm for "gene ordering". Given a matrix of gene expression data values, the task is to find a permutation of the gene names list such that genes with similar expression patterns should be relatively close in the permutation. The algorithm is based on a combined approach that integrates a constructive heuristic with evolutionary and Tabu Search techniques in a single methodology. To evaluate the benefits of this method, we compared our results with the current outputs provided by several widely used algorithms in functional genomics. We also compared the results with our own hierarchical clustering method when used in isolation. We show that the use of images, corrupted with known levels of noise, helps to illustrate some aspects of the performance of the algorithms and provide a complementary benchmark for the analysis. The use of these images, with known high-quality solutions, facilitates in some cases the assessment of the methods and helps the software development, validation and reproducibility of results. We also propose two quantitative measures of performance for gene ordering. Using these measures, we make a comparison with probably the most used algorithm (due to Eisen and collaborators, PNAS 1998) using a microarray dataset available on the public domain (the complete yeast cell cycle dataset).

Algorithms↗

PubMed2001

S-TREE: self-organizing trees for data clustering and online vector quantization.

This paper introduces S-TREE (Self-Organizing Tree), a family of models that use unsupervised learning to construct hierarchical representations of data and online tree-structured vector quantizers. The S-TREE1 model, which features a new tree-building algorithm, can be implemented with various cost functions. An alternative implementation, S-TREE2, which uses a new double-path search procedure, is also developed. The performance of the S-TREE algorithms is illustrated with data clustering and vector quantization examples, including a Gauss-Markov source benchmark and an image compression application. S-TREE performance on these tasks is compared with the standard tree-structured vector quantizer (TSVQ) and the generalized Lloyd algorithm (GLA). The image reconstruction quality with S-TREE2 approaches that of GLA while taking less than 10% of computer time. S-TREE1 and S-TREE2 also compare favorably with the standard TSVQ in both the time needed to create the codebook and the quality of image reconstruction.

Algorithms↗

PubMed2005

On pattern matching of X-ray powder diffraction data.

We introduce a novel pattern matching algorithm optimized for X-ray powder diffraction (XRPD) data and useful for data from other types of analytical techniques (e.g., Raman, IR). The algorithm is based on hierarchical clustering with a similarity metric that compares peak positions using the full peak profile. It includes heuristics developed from years of experience manually matching XRPD data, and preprocessing algorithms that reduce the effects of common problems associated with XRPD (e.g., preferred orientation and poor particle statistics). This algorithm can find immediate application in automated polymorph screening and salt selection, common tasks in the development of pharmaceuticals.

Algorithms↗

PubMedDate not supplied

Identifying conserved gene clusters in the presence of homology families.

The study of conserved gene clusters is important for understanding the forces behind genome organization and evolution, as well as the function of individual genes or gene groups. In this paper, we present a new model and algorithm for identifying conserved gene clusters from pairwise genome comparison. This generalizes a recent model called "gene teams." A gene team is a set of genes that appear homologously in two or more species, possibly in a different order yet with the distance of adjacent genes in the team for each chromosome always no more than a certain threshold. We remove the constraint in the original model that each gene must have a unique occurrence in each chromosome and thus allow the analysis on complex prokaryotic or eukaryotic genomes with extensive paralogs. Our algorithm analyzes a pair of chromosomes in O(mn) time and uses O(m+n) space, where m and n are the number of genes in the respective chromosomes. We demonstrate the utility of our methods by studying two bacterial genomes, E. coli K-12 and B. subtilis. Many of the teams identified by our algorithm correlate with documented E. coli operons, while several others match predicted operons, previously suggested by computational techniques. Our implementation and data are publicly available at euler.slu.edu/ approximately goldwasser/homologyteams/.

Bacillus subtilis↗

PubMed2004

Prediction of cardiovascular risk in hemodialysis patients by data mining.

OBJECTIVES: The objective of this work was to contribute to the development, validation and application of data mining methods for prediction in decision support systems in medicine. The particular focus was on the prediction of cardiovascular risk factors in hemodialysis patients, specifically the interventricular septum (IVS) thickness of the heart of individual patients as an important quantitative indicator to diagnose left ventricular hypertrophy. The work was based on data from 63 long-term hemodialysis patients of the KfH Dialysis Centre in Jena, Germany. METHODS: The approach applied is based on data mining methods and involves four major steps: data based clustering, cluster based rule extraction, rulebase construction and cluster and rule based prediction. The methods employed include crisp and fuzzy algorithms. At each step, logical and medical validation of results was carried out. Different sets of randomly selected patient data were used to train, test and optimize the clusterbases and rulebases for prediction. RESULTS: Using the best clusterbase/rulebase combination designed, the IVS thickness cluster ('small' or 'large') was predicted correctly for 30 of the 35 patients with known IVS values in the training data set; no patient was predicted incorrectly and 5 were parity predicted. For the test data set, 4 of the 6 patients with known IVS values were predicted correctly, no patient incorrectly and 2 parity. These results did not substantially differ from those obtained using the second best clusterbase/rulebase combination which was finally recommended for use based on further performance criteria. The prediction of the IVS thickness clusters of the 22 patients with unknown IVS values also yielded good results that were (and could only be) validated by a medical individual risk assessment of these patients. CONCLUSIONS: The approach applied proved successful for the cluster and rule based prediction of a quantitative variable, such as IVS thickness, for individual patients from other variables relevant to the problem. The results obtained demonstrate the high potential of the approach and the methods developed and validated to support decision-making in hemodialysis and other fields of medicine by individual risk prediction.

Algorithms↗

PubMed1992

An approach based on two-dimensional graph theory for structural cluster detection and its histopathological application.

An approach based on graph theory is described for detecting clusters of cells in tissue specimens (two-dimensional space). With a set of discrete basic elements (cell nuclei) having several measurable features (area, surface, main and minor axis of best-fitting ellipses) a graph is defined as having attributes associated with edges. Different minimum spanning trees (MSTs) can be constructed using different weight functions on the attributes (attributed MST). Analysis of the MST and of an attributed MST by use of a decomposition function allows detection of image areas with similar local properties. These clusters, which are then clusters of the tree, describe, for example, partial growth in different directions in a case of a human fibrosarcoma assuming that tumour cell nuclei are homogeneous with respect to their configuration and size. The model allows the separation of clusters of tumour cells growing in different directions and the approximation of the different growth angles. This decomposition also allows us to create new (higher) orders of structure (cluster tree).

Algorithms↗

PubMed2005

A robust algorithm for ratio estimation in two-color microarray experiments.

The reliability of the algorithms for ratio estimation in two-color microarray image analysis is very important, as these ratios build up the primary source of information for the subsequent analytical procedures (normalization, clustering, classification, etc). Although various algorithms already exist, there is still a need to develop procedures having higher levels of accuracy and robustness. We present a statistical procedure for the detection and removal of aberrant pixels in two-color microarray images. It is based on a linear regression approach, assuming reasonably high level of correlation between the two color channels. This procedure ensures more robust ratio estimation for the spots in both linear regression and traditional segmentation algorithms. The developed algorithms have been evaluated using simulated artificial images and experimental images of different designs. A demonstration version of the software can be downloaded from http://bioinfo.curie.fr/projects/maia/.

Algorithms↗

PubMed2005

Co-occurrence patterns of some small-bodied freshwater fishes in southwestern France: implications for fish conservation and environmental management.

We assessed the influence of environmental variables (elevation, stream order, distance from source, catchment area, slope, stream width, and fish species richness) on the co-occurrence patterns of the minnow, the stone loach, and the gudgeon at the stream system scale. A total of 474 sites were classified according to the seven variables using the Self-Organizing Map (neural network), and three clusters were detected (k-means algorithm). The frequency of the various fish co-occurrence patterns was calculated for each cluster, and general linear modeling was used to specify the conditions that predict the occurrence of each species. Piedmont streams were more likely to support coexisting gudgeon and minnow populations because of higher probabilities of occurrence for both species. The higher co-occurrence frequency for the three species together in headwater streams resulted from lower occurrence frequencies in gudgeon and minnow. Focusing on areas that favor the co-occurrence of species may enhance the effectiveness of conservation projects.

Animals↗

PubMed2000

Cluster, function and promoter: analysis of yeast expression array.

Gene clusters could be derived based on expression profiles, function categorization and promoter regions. To obtain thorough understanding of gene expression and regulation, the three aspects should be combined in an organic way. In this study, we explored the possible ways to analyze the large-scale gene expression data. Three approaches were used to analyze yeast temporal expression data: 1) start from clustering on the expression profiles followed by function categorization and promoter analysis, 2) start from function categorization followed by clustering on expression profiles and promoter analysis, and 3) start from clustering on the promoter region followed by clustering on expression profiles. For clustering analysis on the time-series data, we developed a largest-first algorithm, which provide a mechanism for quality control on clusters. For promoter analysis, we developed a core-extension algorithm.

Algorithms↗