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 1,495 records · Page 83Linked to original sources

PubMedDate not supplied

Measuring stage of change for assessing readiness to increase fruit and vegetable intake among 18- to 24-year-olds.

PURPOSE: To examine the appropriateness of the five-stage schema developed for addictive behaviors when applied to nonaddictive behaviors such as fruit and vegetable consumption. DESIGN: A cross-sectional mail survey was conducted. SETTING: Self-administered questionnaires were mailed to the home addresses of respondents recruited in 10 states and returned upon completion. SUBJECTS: In all, 116 male and 185 female young adults (response rate = 70%) completed the questionnaires. Most respondents were Caucasian (74.4%) and college students (62.2%). MEASURES: Statistical analyses included analysis of variance with Scheffe's test, Pearson correlation and Cronbach's alpha coefficients, factor analysis, and cluster analysis. RESULTS: Staging algorithms for fruits and vegetables were developed according to the five-stage schema, and their construct validity was verified by the incremental trend of food intakes. The algorithms identified more than 80% of the subjects as precontemplators, preparers, or maintainers for both food groups. This distribution pattern agreed well with the three-factor solution generated from the stages of change questionnaire, for which internal reliability and validity were confirmed. Seven cluster profiles of the respondents were derived based on the three generated factors and appeared to represent stage subgroupings of those factors. The two food groups shared marked similarities regarding the profiling of the clusters, which was validated by comparing food intake. The results, however, cannot be generalized because of limitations involved in the current sample. CONCLUSIONS: The original five stages of change and traditional staging measures may need to be adapted and/or extended for dietary change.

Adolescent↗

PubMed2003

Validating cluster size inference: random field and permutation methods.

Cluster size tests used in analyses of brain images can have more sensitivity compared to intensity based tests. The random field (RF) theory has been widely used in implementation of such tests, however the behavior of such tests is not well understood, especially when the RF assumptions are in doubt. In this paper, we carried out a simulation study of cluster size tests under varying smoothness, thresholds, and degrees of freedom, comparing RF performance to that of the permutation test, which is known to be exact. For Gaussian images, we find that the RF methods are generally conservative, especially for low smoothness and low threshold. For t images, the RF tests are found to be conservative at lower thresholds and do not perform well unless the threshold is high and images are sufficiently smooth. The permutation test performs well for any settings though the discreteness in cluster size must be accounted for. We make specific recommendations on when permutation tests are to be preferred to RF tests.

Algorithms↗

PubMed2004

Convergence condition and efficient implementation of the fuzzy curve-tracing (FCT) algorithm.

The fuzzy curve-tracing (FCT) algorithm can be used to extract a smooth curve from unordered noisy data. In this paper, we analyze the convergence property of the algorithm based on the diagonal dominance requirement of the matrix used in the clustering procedure and prove that the algorithm is guaranteed to converge if the weighting coefficient for the smoothness constraint is chosen properly. Based on the convergence condition, we develop several methods for fast and reliable implementation of the algorithm. We show that the algorithm can be initialized with a user-defined curve in many cases, that a multiresolution clustering based approach and an image down-sampling scheme can be used to improve the algorithm stability and speed and that two types of traps can be removed to correct the mistakes in curve tracing. We demonstrate several advantages of our algorithm over the commonly used snake models for boundary detection and several methods for principle curve extraction.

Journal Article↗

PubMed2003

Gene teams: a new formalization of gene clusters for comparative genomics.

This paper describes an efficient algorithm based on a new concept called gene team for detecting conserved gene clusters among an arbitrary number of chromosomes. Within the clusters, neither the order of the genes nor their orientation need be conserved. In addition, insertion of foreign genes within the clusters are permitted to a user-defined extent. This algorithm has been implemented in a publicly available TEAM software that proves to be an efficient tool for systematic searches of conserved gene clusters. Examples of actual biological results are provided. The software is downloadable from http://www-igm.univ-mlv.fr/ approximately raffinot/geneteam.html.

Algorithms↗

PubMed2006

Incorporating biological knowledge into distance-based clustering analysis of microarray gene expression data.

MOTIVATION: Because co-expressed genes are likely to share the same biological function, cluster analysis of gene expression profiles has been applied for gene function discovery. Most existing clustering methods ignore known gene functions in the process of clustering. RESULTS: To take advantage of accumulating gene functional annotations, we propose incorporating known gene functions into a new distance metric, which shrinks a gene expression-based distance towards 0 if and only if the two genes share a common gene function. A two-step procedure is used. First, the shrinkage distance metric is used in any distance-based clustering method, e.g. K-medoids or hierarchical clustering, to cluster the genes with known functions. Second, while keeping the clustering results from the first step for the genes with known functions, the expression-based distance metric is used to cluster the remaining genes of unknown function, assigning each of them to either one of the clusters obtained in the first step or some new clusters. A simulation study and an application to gene function prediction for the yeast demonstrate the advantage of our proposal over the standard method.

Algorithms↗

PubMed2006

Rapid identification of coagulase-negative staphylococci by Fourier transform infrared spectroscopy.

Coagulase-negative staphylococci (CNS), frequently associated with both community-acquired and nosocomial bloodstream infections, must be distinguished from Staphylococcus aureus for clinical purposes. Conventional methods are too laborious and time-consuming and often lack sensitivity to CNS. Fourier transform infrared (FTIR) spectroscopy combined with the use of a universal growth medium (Que-Bact Universal Medium No. 2) and chemometrics was evaluated for its potential as a rapid and simple clinical tool for making this distinction. FTIR spectra of 11 methicillin-sensitive and 11 methicillin-resistant CNS isolates as well as 25 methicillin-sensitive, 47 methicillin-resistant, 34 borderline oxacillin-resistant and 35 glycopeptide intermediate S. aureus isolates were obtained from dried films of stationary-phase cells grown on the universal medium. Principal component analysis (PCA), self-organizing maps, and the K-nearest neighbor algorithm were employed to cluster the different phenotypes based on similarity of their FTIR spectra. PCA of the first-derivative normalized spectral data from a single narrow region (2888-2868 cm(-1)) yielded complete differentiation of CNS from both methicillin-sensitive and methicillin-resistant S. aureus. The rate of correct classification was somewhat reduced, from 100% to 90%, after inclusion of borderline oxacillin-resistant and glycopeptide intermediate S. aureus strains in the data set. Differentiation based on the data in broader spectral regions was much less reliable. The results of this study indicate that with proper spectral region selection, FTIR spectroscopy and cluster analysis may provide a simple and accurate means of CNS species identification.

Algorithms↗

PubMed2006

Implementation and performance evaluation of reconstruction algorithms on graphics processors.

The high-throughput needs in electron tomography and in single particle analysis have driven the parallel implementation of several reconstruction algorithms and software packages on computing clusters. Here, we report on the implementation of popular reconstruction algorithms as weighted backprojection, simultaneous iterative reconstruction technique (SIRT) and simultaneous algebraic reconstruction technique (SART) on common graphics processors (GPUs). The speed gain achieved on the GPUs is in the order of sixty (60x) to eighty (80x) times, compared to the performance of a single central processing unit (CPU), which is comparable to the acceleration achieved on a medium-range computing cluster. This acceleration of the reconstruction is caused by the highly specialized architecture of the GPU. Further, we show that the quality of the reconstruction on the GPU is comparable to the CPU. We present detailed flow-chart diagrams of the implementation. The reconstruction software does not require special hardware apart from the commercially available graphics cards and could be easily integrated in software packages like SPIDER, XMIPP, TOM-Package and others.

Algorithms↗

PubMed1996

Prediction of HIV peptide epitopes by a novel algorithm.

Identification of promiscuous or multideterminant T cell epitopes is essential for HIV vaccine development, however, current methods for T cell epitope identification are both cost intensive and labor intensive. We have developed a computer-driven algorithm, named EpiMer, which searches protein amino acid sequences for putative MHC class I- and/or class II-restricted T cell epitopes. This algorithm identifies peptides that contain multiple MHC-binding motifs from protein sequences. To evaluate the predictive power of EpiMer, the amino acid sequences of the HIV-1 proteins nef, gp160, gag p55, and tat were searched for regions of MHC-binding motif clustering. We assessed the algorithm's predictive power by comparing the EpiMer-predicted peptide epitopes to T cell epitopes that have been published in the literature. The EpiMer method of T cell epitope identification was compared to the standard method of synthesizing short, overlapping peptides and testing them for immunogenicity (overlapping peptide method), and to an alternate algorithm that has been used to identify putative T cell epitopes from primary structure (AMPHI). For the four HIV-1 proteins analyzed, the in vitro testing of EpiMer peptides for immunogenicity would have required the synthesis of fewer total peptides than either AMPHI or the overlapping peptide method. The EpiMer algorithm proved to be more efficient and more sensitive per amino acid than both the overlapping peptide method and AMPHI. The EpiMer predictions for these four HIV proteins are described. Since EpiMer-predicted peptides have the potential to bind to multiple MHC alleles, they are strong candidates for inclusion in a synthetic HIV vaccine.

Algorithms↗

PubMed2002

Application of high-throughput computing in bioinformatics.

We describe the challenges faced when developing a Linux/PC-based cluster to apply bioinformatics algorithms to the rapidly increasing raw genomics data available. The calculations, which take around two months to complete, result in a powerful resource that can be used for data mining--most obviously for the human genome. Our current infrastructure consists of a 1314 node cluster with 1734 processors supporting both production and research. This paper highlights the problems in achieving high data throughput with such systems and shows that raw computer power is only one component of a complex problem.

Computational Biology↗

PubMed2004

Identifying time-lagged gene clusters using gene expression data.

MOTIVATION: Analysis of gene expression data can provide insights into the time-lagged co-regulation of genes/gene clusters. However, existing methods such as the Event Method and the Edge Detection Method are inefficient as they compare only two genes at a time. More importantly, they neglect some important information due to their scoring criterian. In this paper, we propose an efficient algorithm to identify time-lagged co-regulated gene clusters. The algorithm facilitates localized comparison and processes several genes simultaneously to generate detailed and complete time-lagged information for genes/gene clusters. RESULTS: We experimented with the time-series Yeast gene dataset and compared our algorithm with the Event Method. Our results show that our algorithm is not only efficient, but also delivers more reliable and detailed information on time-lagged co-regulation between genes/gene clusters. AVAILABILITY: The software is available upon request. CONTACT: jiliping@comp.nus.edu.sg SUPPLEMENTARY INFORMATION: Supplementary tables and figures for this paper can be found at http://www.comp.nus.edu.sg/~jiliping/p2.htm.

Cluster Analysis↗

PubMed2003

Identification and quantification of disease-related gene clusters.

MOTIVATION: DNA microarray technology and the completion of human and mouse genome sequencing programs are now offering new avenues for the investigation of complex genetic diseases. In particular, this makes possible the study of the spatial distribution of disease-related genes within the genome. We report on the first systematic search for clustering of genes associated with a polygenic autoimmune disease. RESULTS: Using a set of cDNA microarray chip experiments in two mouse models of rheumatoid arthritis, we have identified approximately 200 genes based on their expression in inflamed joints and mapped them into the genome. We compute the spatial autocorrelation function of the selected genes and find that they tend to cluster over scales of a few megabase pairs. We then identify significant gene clusters using a friends-of-friends algorithm. This approach should aid in discovering functionally related gene clusters in the mammalian genome.

Algorithms↗

PubMed1997

Post-processing of BLAST results using databases of clustered sequences.

MOTIVATION: When evaluating the results of a sequence similarity search, there are many situations where it can be useful to determine whether sequences appearing in the results share some distinguishing characteristic. Such dependencies between database entries are often not readily identifiable, but can yield important new insights into the biological function of a gene or protein. RESULTS: We have developed a program called CBLAST that sorts the results of a BLAST sequence similarity search according to sequence membership in user-defined 'clusters' of sequences. To demonstrate the utility of this application, we have constructed two cluster databases. The first describes clusters of nucleotide sequences representing the same gene, as documented in the UNIGENE database, and the second describes clusters of protein sequences which are members of the protein families documented in the PROSITE database. Cluster databases and the CBLAST post-processor provide an efficient mechanism for identifying and exploring relationships and dependencies between new sequences and database entries.

Algorithms↗

PubMed2004

A fast sequential image fractal coding approach based on optimal fuzzy clustering.

To reduce the coding time of the conventional method, a fast sequential image fractal compression algorithm was proposed on the basis of the principle of optimal fuzzy clustering (OFC) for an unsupervised sample set with the category number settled by the algorithm itself. We utilized the cost function defined by the OFC algorithm to obtain the best category number corresponding to the minimum value of the function. Firstly the Linde-Buzo-Gray (LBG) algorithm was realized to acquire a rough cluster of the domain pool. Then the optimal category number was obtained by implementing our algorithm with small computational cost. Finally the more precise category was gained and the detail of the reconstructed image efficiently preserved. As a global optimal algorithm, OFC not only helps LBG eliminate the local minima, but also effectively compensates for the arbitrary interference in hard clustering problem. Soft clustering of the domain blocks allows classified searches instead of global ones and takes less coding time, and therefore clearly outperforms to the classic method relying on reduction of the size of the domain pool by classification. In computer simulation, OFC-based algorithm for the fractal coding scheme achieved excellent performance. For some standard and sequential medical images, the results denoted that the encoding speed was improved by about 5 folds without affecting the signal-to-noise ratio and compression ratio, and the quality of the reconstructed image could be better retained.

Algorithms↗

PubMed2006

Validation and functional annotation of expression-based clusters based on gene ontology.

BACKGROUND: The biological interpretation of large-scale gene expression data is one of the paramount challenges in current bioinformatics. In particular, placing the results in the context of other available functional genomics data, such as existing bio-ontologies, has already provided substantial improvement for detecting and categorizing genes of interest. One common approach is to look for functional annotations that are significantly enriched within a group or cluster of genes, as compared to a reference group. RESULTS: In this work, we suggest the information-theoretic concept of mutual information to investigate the relationship between groups of genes, as given by data-driven clustering, and their respective functional categories. Drawing upon related approaches (Gibbons and Roth, Genome Research 12:1574-1581, 2002), we seek to quantify to what extent individual attributes are sufficient to characterize a given group or cluster of genes. CONCLUSION: We show that the mutual information provides a systematic framework to assess the relationship between groups or clusters of genes and their functional annotations in a quantitative way. Within this framework, the mutual information allows us to address and incorporate several important issues, such as the interdependence of functional annotations and combinatorial combinations of attributes. It thus supplements and extends the conventional search for overrepresented attributes within a group or cluster of genes. In particular taking combinations of attributes into account, the mutual information opens the way to uncover specific functional descriptions of a group of genes or clustering result. All datasets and functional annotations used in this study are publicly available. All scripts used in the analysis are provided as additional files.

Algorithms↗

PubMed2002

Assessment of the parallelization approach of d2_cluster for high-performance sequence clustering.

The exponential increase in expressed sequence tag (EST) sequence data amplifies the computational cost of clustering sequences such that new algorithms are required to analyze data at a greater rate. We have parallelized d2_cluster on a SGI Origin 2000 multiprocessor and observed a speedup of approximately 100x on 126 processors when processing a 15,876 EST dataset. The parallelized d2_cluster code is obtainable from the SANBI website (http://www.sanbi.ac.za/CODES).

Journal Article↗

PubMed2000

Minimizing stochastic complexity using local search and GLA with applications to classification of bacteria.

In this paper, we compare the performance of two iterative clustering methods when applied to an extensive data set describing strains of the bacterial family Enterobacteriaceae. In both methods, the classification (i.e. the number of classes and the partitioning) is determined by minimizing stochastic complexity. The first method performs the minimization by repeated application of the generalized Lloyd algorithm (GLA). The second method uses an optimization technique known as local search (LS). The method modifies the current solution by making global changes to the class structure and it, then, performs local fine-tuning to find a local optimum. It is observed that if we fix the number of classes, the LS finds a classification with a lower stochastic complexity value than GLA. In addition, the variance of the solutions is much smaller for the LS due to its more systematic method of searching. Overall, the two algorithms produce similar classifications but they merge certain natural classes with microbiological relevance in different ways.

Algorithms↗

PubMed2006

On weighting clustering.

Recent papers and patents in iterative unsupervised learning have emphasized a new trend in clustering. It basically consists of penalizing solutions via weights on the instance points, somehow making clustering move toward the hardest points to cluster. The motivations come principally from an analogy with powerful supervised classification methods known as boosting algorithms. However, interest in this analogy has so far been mainly borne out from experimental studies only. This paper is, to the best of our knowledge, the first attempt at its formalization. More precisely, we handle clustering as a constrained minimization of a Bregman divergence. Weight modifications rely on the local variations of the expected complete log-likelihoods. Theoretical results show benefits resembling those of boosting algorithms and bring modified (weighted) versions of clustering algorithms such as k-means, fuzzy c-means, Expectation Maximization (EM), and k-harmonic means. Experiments are provided for all these algorithms, with a readily available code. They display the advantages that subtle data reweighting may bring to clustering.

Algorithms↗

PubMed2004

Theoretical analysis of microdosimetric spectra and cluster formation for 103Pd and 125I photon emitters.

In this work we have compared 125I or 103Pd from a microdosimetric point of view. The photon spectra at different positions around the seeds have first been calculated using EGSnrc Monte Carlo (MC) code. These photon spectra are used as input for the event-by-event MC code TRION to calculate the microdosimetric lineal energy (y) distribution for each isotope. The microdosimetric dose average lineal energy, yD, calculated in a sphere of 1 microm is 3.5 keV microm(-1) for 125I and 4.0 keV microm(-1) for 103Pd, agreeing well with values reported in the literature. yD in a 1 microm sphere diminishes slightly with the distance from the seed for 103Pd. This is due to the spectral hardening caused by the presence of a gamma-ray of 357.5 keV in the initial spectrum of 103Pd. In parallel with the calculation of the microdosimetric spectra, we have analysed the distribution of the size of the energy deposition clusters generated by these low energy photons in structures of 2 and 10 nm of radius. Due to Compton interactions, the fraction of very low energy electrons (<5 keV) generated by 125I photons is 51%, whereas it is only 27% for 103Pd. As these electrons deposit their energy very locally, the pattern of energy depositions contains more clusters of a few nm of radius for 125I than for 103Pd; the mean cluster orders are respectively 3.3 and 3.0 for 10 nm clusters. This is in opposition with the prediction based on the microdosimetric spectrum and the parameter yD and could be of importance for the damage to the cells.

Algorithms↗