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 1,027 records · Page 57Linked to original sources

PubMed1998

Computational space reduction and parallelization of a new clustering approach for large groups of sequences.

MOTIVATION: The explosive growth of the biological sequences databases stimulated by genome projects has modified the framework of several applications in the biological sequence analysis area. In most cases, this new scenario is characterized by studies on large sets of sequences, suggesting the need for effective and automatic methods for their clustering. A more effective clustering of the database could be followed by the application of common family analysis schemes to the groups so formed. RESULTS: In this work, we present a new strategy to reduce the computational cost associated with the clustering of large sets of sequences which are expected to contain several families. The strategy is based on the grouping of the sequences into families by using a dynamic threshold on a pairwise sequence similarity criterion. Routine clustering of large data sets can now be done very efficiently. The method developed here achieves a computational space reduction of about an order of magnitude over more traditional ones of all-versus-all comparisons. The outcome of this approach produces family groupings that reproduce closely already accepted biological results. Our work includes a parallel implementation for distributed memory multiprocessors with a dynamic scheduling strategy for performance optimization. AVAILABILITY: By anonymous ftp at ftp.ac.uma.es (/pub/ots/pCluster directory), or from our Web site http://www.cnb. uam.es/www/software/software_index.html CONTACT: ots@ac.uma.es

Algorithms↗

PubMedDate not supplied

[An assessment of the clinical status of patients with alveolar echinococcosis by using statistical processing methods. 1. An assessment of the stage of the disease by cluster analysis].

The clinical condition was assessed in 44 patients with alveolar echinococcosis, using methods of mathematical analysis of 10 basic routine laboratory parameters reflecting protein-forming liver function and immune disturbances. The data were processed on a "NOVA-2" [correction of "HOBA-2"] computer using Fortran software on taxonomy 7 classes of conditions were determined, which reflect the progressing severity of the pathological process and correspond to three stages of the disease: uncomplicated, complicated and terminal. A high clinical value of the method has been demonstrated for the evaluation of the patient's condition and determination of the treatment tactics as well as its importance for the creation of a clinical data bank in the invasion focus.

Algorithms↗

PubMed2005

A structure-based expert model of the ICD-10 mental disorders.

BACKGROUND: Many authors have highlighted that a classificatory system such as the ICD-10 or the DSM-IV should attempt to integrate categorical as well as dimensional aspects rather than relying only on a categorical distinction. METHODS: In the current study, a method is presented that allows both perspectives to be considered. Based on their clinical experience, 20 therapists were asked to rate a selection of ICD-10 mental disorders in terms of their similarity. The resulting data were processed by nonmetric multidimensional scaling on individual basis as well as mean basis and a cluster analysis. RESULTS: From a categorical point of view, the resulting mean cognitive map and the cluster analytic groups showed a good concordance with the ICD-10 F groups. Additionally, analogies to the critical observations with regard to the classification systems could be observed. From the dimensional point of view, there were two dimensions found (cognition and psychosis) which showed high correlations in the mean map. CONCLUSIONS: The method presented in this study offers an opportunity to account for dimensional as well as categorical aspects of classification by exploring the structure of the ICD-10 mental disorders directly from the point of view of therapists and thus creating the basis for an expert model.

Algorithms↗

PubMed2005

Mixture modeling with pairwise, instance-level class constraints.

The goal of semisupervised clustering/mixture modeling is to learn the underlying groups comprising a given data set when there is also some form of instance-level supervision available, usually in the form of labels or pairwise sample constraints. Most prior work with constraints assumes the number of classes is known, with each learned cluster assumed to be a class and, hence, subject to the given class constraints. When the number of classes is unknown or when the one-cluster-per-class assumption is not valid, the use of constraints may actually be deleterious to learning the ground-truth data groups. We address this by (1) allowing allocation of multiple mixture components to individual classes and (2) estimating both the number of components and the number of classes. We also address new class discovery, with components void of constraints treated as putative unknown classes. For both real-world and synthetic data, our method is shown to accurately estimate the number of classes and to give favorable comparison with the recent approach of Shental, Bar-Hillel, Hertz, and Weinshall (2003).

Algorithms↗

PubMed2001

Three-dimensional computer-aided diagnosis scheme for detection of colonic polyps.

We have developed a three-dimensional (3-D) computer-aided diagnosis scheme for automated detection of colonic polyps in computed tomography (CT) colonographic data sets, and assessed its performance based on colonoscopy as the gold standard. In this scheme, a thick region encompassing the entire colonic wall is extracted from an isotropic volume reconstructed from the CT images in CT colonography. Polyp candidates are detected by first computing of 3-D geometric features that characterize polyps, folds, and colonic walls at each voxel in the extracted colon, and then segmenting of connected components corresponding to suspicious regions by hysteresis thresholding based on these geometric features. We apply fuzzy clustering to these connected components to obtain the polyp candidates. False-positive (FP) detections are then reduced by computation of several 3-D volumetric features characterizing the internal structures of the polyp candidates, followed by the application of discriminant analysis to the feature space generated by these volumetric features. The locations of the polyps detected by our computerized method were compared to the gold standard of conventional colonoscopy. The performance was evaluated based on 43 clinical cases, including 12 polyps determined by colonoscopy. Our computerized scheme was shown to have the potential to detect polyps in CT colonography with a clinically acceptable high sensitivity and a low FP rate.

Algorithms↗

PubMed2006

Multiscale spatial Monte Carlo simulations: multigriding, computational singular perturbation, and hierarchical stochastic closures.

Monte Carlo (MC) simulation of most spatially distributed systems is plagued by several problems, namely, execution of one process at a time, large separation of time scales of various processes, and large length scales. Recently, a coarse-grained Monte Carlo (CGMC) method was introduced that can capture large length scales at reasonable computational times. An inherent assumption in this CGMC method revolves around a mean-field closure invoked in each coarse cell that is inaccurate for short-ranged interactions. Two new approaches are explored to improve upon this closure. The first employs the local quasichemical approximation, which is applicable to first nearest-neighbor interactions. The second, termed multiscale CGMC method, employs singular perturbation ideas on multiple grids to capture the entire cluster probability distribution function via short microscopic MC simulations on small, fine-grid lattices by taking advantage of the time scale separation of multiple processes. Computational strategies for coupling the fast process at small length scales (fine grid) with the slow processes at large length scales (coarse grid) are discussed. Finally, the binomial tau-leap method is combined with the multiscale CGMC method to execute multiple processes over the entire lattice and provide additional computational acceleration. Numerical simulations demonstrate that in the presence of fast diffusion and slow adsorption and desorption processes the two new approaches provide more accurate solutions in comparison to the previously introduced CGMC method.

Algorithms↗

PubMed2005

Hierarchical clustering of the correlation patterns: new method of domain identification in proteins.

New method of identification of dynamical domains in proteins - Hierarchical Clustering of the Correlation Patterns (HCCP) is proposed. HCCP allows to identify the domains using single three-dimensional structure of the studied proteins and does not require any adjustable parameters that can influence the results. The method is based on hierarchical clustering performed on the matrices of correlation patterns, which are obtained by the transformation of ordinary pairwise correlation matrices. This approach allows to extract additional information from the correlation matrices, which increases reliability of domain identification. It is shown that HCCP is insensitive to small variations of the pairwise correlation matrices. Particularly it produces identical results if the data obtained for the same protein crystallized with different spatial positions of domains are used for analysis. HCCP can utilize correlation matrices obtained by any method such as normal mode or essential dynamics analysis, Gaussian network or anisotropic network models, etc. These features make HCCP an attractive method for domain identification in proteins.

Algorithms↗

PubMed2004

A global test for groups of genes: testing association with a clinical outcome.

MOTIVATION: This paper presents a global test to be used for the analysis of microarray data. Using this test it can be determined whether the global expression pattern of a group of genes is significantly related to some clinical outcome of interest. Groups of genes may be any size from a single gene to all genes on the chip (e.g. known pathways, specific areas of the genome or clusters from a cluster analysis). RESULT: The test allows groups of genes of different size to be compared, because the test gives one p-value for the group, not a p-value for each gene. Researchers can use the test to investigate hypotheses based on theory or past research or to mine gene ontology databases for interesting pathways. Multiple testing problems do not occur unless many groups are tested. Special attention is given to visualizations of the test result, focussing on the associations between samples and showing the impact of individual genes on the test result. AVAILABILITY: An R-package globaltest is available from http://www.bioconductor.org

Algorithms↗

PubMedDate not supplied

Identifying biologically active compound classes using phenotypic screening data and sampling statistics.

Scoring the activity of compounds in phenotypic high-throughput assays presents a unique challenge because of the limited resolution and inherent measurement error of these assays. Techniques that leverage the structural similarity of compounds within an assay can be used to improve the hit-recovery rate from screening data. A technique is presented that uses clustering and sampling statistics to predict likely compound activity by scoring entire structural classes. A set of phenotypic assays performed against a commercially available compound library was used as a test set. Using the class-scoring technique, the resultant activity prediction scores were more reproducible than individual assay measurements, and class scoring recovered known active compounds more efficiently than individual assay measurements because class scoring had fewer false positives. Known biologically active compounds were recovered 87% of the time using class scores, suggesting a low false-negative rate that compared well to individual assay measurements. In addition, many weak and potentially novel classes of active compounds, overlooked by individual assay measurements, were suggested.

Actins↗

PubMed2005

Color clustering and learning for image segmentation based on neural networks.

An image segmentation system is proposed for the segmentation of color image based on neural networks. In order to measure the color difference properly, image colors are represented in a modified L*u* v* color space. The segmentation system comprises unsupervised segmentation and supervised segmentation. The unsupervised segmentation is achieved by a two-level approach, i.e., color reduction and color clustering. In color reduction, image colors are projected into a small set of prototypes using self-organizing map (SOM) learning. In color clustering, simulated annealing (SA) seeks the optimal clusters from SOM prototypes. This two-level approach takes the advantages of SOM and SA, which can achieve the near-optimal segmentation with a low computational cost. The supervised segmentation involves color learning and pixel classification. In color learning, color prototype is defined to represent a spherical region in color space. A procedure of hierarchical prototype learning (HPL) is used to generate the different sizes of color prototypes from the sample of object colors. These color prototypes provide a good estimate for object colors. The image pixels are classified by the matching of color prototypes. The experimental results show that the system has the desired ability for the segmentation of color image in a variety of vision tasks.

Algorithms↗

PubMed2004

ESTminer: a Web interface for mining EST contig and cluster databases.

UNLABELLED: ESTminer is a Web application and database schema for interactive mining of expressed sequence tag (EST) contig and cluster datasets. The Web interface contains a query frame that allows the selection of contigs/clusters with specific cDNA library makeup or a threshold number of members. The results are displayed as color-coded tree nodes, where the color indicates the fractional size of each cDNA library component. The nodes are expandable, revealing library statistics as well as EST or contig members, with links to sequence data, GenBank records or user configurable links. Also, the interface allows 'queries within queries' where the result set of a query is further filtered by the subsequent query. AVAILABILITY: ESTminer is implemented in Java/JSP and the package, including MySQL and Oracle schema creation scripts, is available from http://cggc.agtec.uga.edu/Data/download.asp CONTACT: agingle@uga.edu.

Algorithms↗

PubMed2003

Combinatorial image analysis of DNA microarray features.

MOTIVATION: DNA and protein microarrays have become an established leading-edge technology for large-scale analysis of gene and protein content and activity. Contact-printed microarrays has emerged as a relatively simple and cost effective method of choice but its reliability is especially susceptible to quality of pixel information obtained from digital scans of spotted features in the microarray image. RESULTS: We address the statistical computation requirements for optimizing data acquisition and processing of digital scans. We consider the use of median filters to reduce noise levels in images and top-hat filters to correct for trends in background values. We also consider, as alternative estimators of spot intensity, discs of fixed radius, proportions of histograms and k-means clustering, either with or without a square-root intensity transformation and background subtraction. We identify, using combinatoric procedures, optimal filter and estimator parameters, in achieving consistency among the replicates of a gene on each microarray. Our results, using test data from microarrays of HCMV, indicate that a highly effective approach for improving reliability and quality of microarray data is to apply a 21 by 21 top-hat filter, then estimate spot intensity as the mean of the largest 20% of pixel values in the target region, after a square-root transformation, and corrected for background, by subtracting the mean of the smallest 70% of pixel values. AVAILABILITY: Fortran90 subroutines implementing these methods are available from the authors, or at http://www.bioss.ac.uk/~chris.

Algorithms↗

PubMed2002

Selecting targets for structural determination by navigating in a graph of protein families.

MOTIVATION: A major goal in structural genomics is to enrich the catalogue of proteins whose 3D structures are known. In an attempt to address this problem we mapped over 10 000 proteins with solved structures onto a graph of all Swissprot protein sequences (release 36, approximately 73 000 proteins) provided by ProtoMap, with the goal of sorting proteins according to their likelihood of belonging to new superfamilies. We hypothesized that proteins within neighbouring clusters tend to share common structural superfamilies or folds. If true, the likelihood of finding new superfamilies increases in clusters that are distal from other solved structures within the graph. RESULTS: We defined an order relation between unsolved proteins according to their 'distance' from solved structures in the graph, and sorted approximately 48 000 proteins. Our list can be partitioned into three groups: approximately 35 000 proteins sharing a cluster with at least one known structure; approximately 6500 proteins in clusters with no solved structure but with neighbouring clusters containing known structures; and a third group contains the rest of the proteins, approximately 6100 (in 1274 clusters). We tested the quality of the order relation using thousands of recently solved structures that were not included when the order was defined. The tests show that our order is significantly better (P-value approximately 10(5)) than a random order. More interestingly, the order within the union of the second and third groups, and the order within the third group alone, perform better than random (P-values: 0.0008 and 0.15, respectively) and are better than alternative orders created using PSI-BLAST. Herein, we present a method for selecting targets to be used in structural genomics projects. AVAILABILITY: List of proteins to be used for targets selection combined with a set of biological filters for narrowing down potential targets is in http://www.protarget.cs.huji.ac.il.

Algorithms↗

PubMed2005

Theoretical fine spectroscopy with symmetry-adapted-cluster configuration-interaction method: outer- and inner-valence ionization spectra of furan, pyrrole, and thiophene.

Theoretical fine spectroscopy has been performed for the valence ionization spectra of furan, pyrrole, and thiophene with the symmetry-adapted-cluster configuration-interaction general-R method. The present method described that the pi(1) state interacts with the pi(3) (-2)pi*, pi(2) (-2)pi*, and pi(2) (-1)pi(3) (-1)pi* shake-up states providing the split peaks and the outer-valence satellites, both of which are in agreement with the experiments. The intensity distributions were analyzed in detail for the inner-valence region. In particular, for furan, theoretical intensities were successfully compared with the intensity measured by the electron momentum spectroscopy. The interactions of the 3b(2) and 5a(1) states with the shake-up states were remarkable for furan and pyrrole, while the 4b(2) state of thiophene had relatively large intensity.

Algorithms↗

PubMed2004

Cluster distance geometry of polypeptide chains.

Distance geometry has been a broadly useful tool for dealing with conformational calculations. Customarily each atom is represented as a point, constraints on the distances between some atoms are obtained from experimental or theoretical sources, and then a random sampling of conformations can be calculated that are consistent with the constraints. Although these methods can be applied to small proteins having on the order of 1000 atoms, for some purposes it is advantageous to view the problem at lower resolution. Here distance geometry is generalized to deal with distances between sets of points. In the end, much of the same techniques produce a sampling of different configurations of these sets of points subject to distance constraints, but now the radii of gyration of the different sets play an important role. A simple example is given of how the packing constraints for polypeptide chains combine with loose distance constraints to give good calculated protein conformers at a very low resolution.

Algorithms↗

PubMed1997

High-frequency FSH and LH pulses in obese menopausal women.

We studied the pulsatile luteinizing hormone (LH) and follicle-stimulating hormone (FSH) secretory patterns, at early or intermediate years of menopause in seven normal women with different degrees of obesity, taking blood samples every minute for 40 min to 2 h. The hormones were assayed with an immunoradiometric assay (IRMA) system, analyzing with the cluster pulse algorithm. All women showed hormone pulses every 8-10 min. In five of them were found periods of discrete pulses with oscillations of high amplitude alternating with periods of pulses of low amplitude. In two cases, the high-frequency oscillatory pattern with low amplitude was found around low mean levels of 22.8 and 25.7 IU/L. The LH oscillatory pattern also had a high frequency, but at a lower level, giving a high FSH/LH ratio. The coincidence index of FSH with LH peaks was 76.6%. We concluded that at menopause, the frequency of FSH and LH secretion increases with a high FSH/LH ratio. Obese menopausal women may have the same high-frequency oscillatory patterns, but at low levels.

Body Mass Index↗

PubMed2006

Independent confirmation of a prognostic gene-expression signature in adult acute myeloid leukemia with a normal karyotype: a Cancer and Leukemia Group B study.

Patients with acute myeloid leukemia (AML) and normal karyotype are classified in an intermediate-risk group, albeit this subset is heterogeneous for clinical outcome. A recent complementary DNA microarray study identified a gene-expression signature that--when used to cluster normal karyotype patients--separated them into 2 prognostically relevant subgroups. We sought the first independent validation of the prognostic value of this signature. Using oligonucleotide microarrays to measure gene expression in samples from uniformly treated adults with karyotypically normal AML, we performed cluster analysis based on the previously identified signature. We also developed a well-defined classification rule using the signature to predict outcome for individual patients. Cluster analysis confirmed the prognostic utility of the signature: patient clusters differed in overall (P = .001) and disease-free (P = .001) survival. The signature-based classifier identified groups with differences in overall (P = .02) and disease-free (P = .05) survival. A strong association of the outcome classifier with the prognostically adverse FLT3 internal tandem duplication (FLT3 ITD) potentially explained the prognostic significance of the signature. However, in the subgroup of patients without FLT3 ITD there was a moderate difference in survival for the classifier-derived groups. Our analysis confirms the applicability of the gene-expression profiling strategy for outcome prediction in cytogenetically normal AML.

Acute Disease↗

PubMed2002

Large-scale prediction of Saccharomyces cerevisiae gene function using overlapping transcriptional clusters.

Genome sequencing has led to the discovery of tens of thousands of potential new genes. Six years after the sequencing of the well-studied yeast Saccharomyces cerevisiae and the discovery that its genome encodes approximately 6,000 predicted proteins, more than 2,000 have not yet been characterized experimentally, and determining their functions seems far from a trivial task. One crucial constraint is the generation of useful hypotheses about protein function. Using a new approach to interpret microarray data, we assign likely cellular functions with confidence values to these new yeast proteins. We perform extensive genome-wide validations of our predictions and offer visualization methods for exploration of the large numbers of functional predictions. We identify potential new members of many existing functional categories including 285 candidate proteins involved in transcription, processing and transport of non-coding RNA molecules. We present experimental validation confirming the involvement of several of these proteins in ribosomal RNA processing. Our methodology can be applied to a variety of genomics data types and organisms.

Algorithms↗