Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “unsupervised clustering”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 541 records · Page 30Linked to original sources
PubMed2003

CNN universal machine as classificaton platform: an art-like clustering algorithm.

Fast and robust classification of feature vectors is a crucial task in a number of real-time systems. A cellular neural/nonlinear network universal machine (CNN-UM) can be very efficient as a feature detector. The next step is to post-process the results for object recognition. This paper shows how a robust classification scheme based on adaptive resonance theory (ART) can be mapped to the CNN-UM. Moreover, this mapping is general enough to include different types of feed-forward neural networks. The designed analogic CNN algorithm is capable of classifying the extracted feature vectors keeping the advantages of the ART networks, such as robust, plastic and fault-tolerant behaviors. An analogic algorithm is presented for unsupervised classification with tunable sensitivity and automatic new class creation. The algorithm is extended for supervised classification. The presented binary feature vector classification is implemented on the existing standard CNN-UM chips for fast classification. The experimental evaluation shows promising performance after 100% accuracy on the training set.

Algorithms↗
PubMed2005

Counting moles automatically from back images.

Density of moles is a strong predictor of malignant melanoma, therefore, enumeration of moles is often an integral part of many studies that look at malignant melanoma. An automatic method of segmenting and counting moles would help standardize studies, compared with manual counting. We have developed an unsupervised algorithm for segmenting and counting moles from two-dimensional color images of the back torso region, as part of a study to evaluate the effectiveness of sunscreen. The method consists of a new variant of mean shift filtering that forms clusters in the image and removes noise, a region growing procedure to select candidates, and a rule-based classifier to identify moles. When this algorithm was compared to an assessment by an expert dermatologist, the algorithm showed a sensitivity rate of 91% and diagnostic accuracy of 90% on the test set, for moles larger than 1.5 mm in diameter.

Algorithms↗
PubMedDate not supplied

Combining unsupervised and supervised artificial neural networks to predictaquatic toxicity.

Most quantitative structure-activity relationship (QSAR) models are linear relationships and significant for only a limited domain of compounds. Here we propose a data-driven approach with a flexible combination of unsupervised and supervised neural networks able to predict the toxicity of a large set of different chemicals while still respecting the QSAR postulates. Since QSAR is applicable only to similar compounds, which have similar biological and physicochemical properties, large numbers of compounds are clustered before building local models, and local models are ensembled to obtain the final result. The approach has been used to develop models to predict the fish toxicity of Pimephales promelas and Tetrahymena pyriformis, a protozoan.

Journal Article↗
PubMed2006

Protein clusters associated with carcinogenesis, histological differentiation and nodal metastasis in esophageal cancer.

We examined the proteomic background of esophageal cancer. We used laser microdissection to obtain tumor tissues from 72 esophageal squamous cell carcinoma cases and adjacent normal tissues in 57 of these cases. The 2D-DIGE generated quantitative expression profiles with 1730 protein spots. Based on the intensity of the protein spots, unsupervised classification distinguished the tumor tissues from their normal counterparts, and subdivided the tumor tissues according to their histological differentiation. We identified 498 protein spots with altered intensity in the tumor tissues, which protein identification by LC-MS/MS showed to correspond to 217 gene products. We also found 41 protein spots that were associated with nodal metastasis, and identified 33 proteins corresponding to the spots, including cancer-associated proteins such as alpha-actinin 4, hnRNP K, periplakin, squamous cell carcinoma antigen 1 and NudC. The identified cancer-associated proteins have been previously reported to be individually involved in a range of cancer types, and our study observed them collectively in a single type of malignancy, esophageal cancer. As the identified proteins are involved in important biological processes such as cytoskeletal/structural organization, transportation, chaperon, oxidoreduction, transcription and signal transduction, they may function in a coordinate manner in carcinogenesis and tumor progression of esophageal cancer.

Carcinoma, Squamous Cell↗
PubMed2005

A bayesian recurrent neural network for unsupervised pattern recognition in large incomplete data sets.

A recurrent neural network, modified to handle highly incomplete training data is described. Unsupervised pattern recognition is demonstrated in the WHO database of adverse drug reactions. Comparison is made to a well established method, AutoClass, and the performances of both methods is investigated on simulated data. The neural network method performs comparably to AutoClass in simulated data, and better than AutoClass in real world data. With its better scaling properties, the neural network is a promising tool for unsupervised pattern recognition in huge databases of incomplete observations.

Algorithms↗
PubMed1998

Can functional regions of proteins be predicted from their coding sequences? The case study of G-protein coupled receptors.

A filter based on a set of unsupervised neural networks trained with a winner-take-all strategy discloses signals along the coding sequences of G-protein coupled receptors. By comparing with the existing experimental data it appears that these signals correlate with putative functional domains of the proteins. After protein alignment within subfamilies, signals cluster in protein regions which, according to the presently available experimental results, are described as possible functional domains of the folded proteins. The mapping procedure reveals characteristic regions in the coding sequences common and/or characteristic of the receptor subtype. This is particularly noticeable for the third cytoplasmic loop, which is likely to be involved in the molecular coupling of all the subfamilies with G-proteins. The results indicate that our mapping can highlight intrinsic representative features of the coding sequences which, in the case of G-protein coupled receptors, are characteristic of protein functional regions and suggest a possible application of the filter for predicting functional determinants in proteins starting from the coding sequence.

Amino Acid Sequence↗
PubMed2004

Microarray gene expression profiling of B-cell chronic lymphocytic leukemia subgroups defined by genomic aberrations and VH mutation status.

PURPOSE: Genomic aberrations and mutational status of the immunoglobulin variable heavy chain (VH) gene have been shown to be among the most important predictors for outcome in patients with B-cell chronic lymphocytic leukemia (B-CLL). In this study, we report on differential gene expression patterns that are characteristic for genetically defined B-CLL subtypes. MATERIALS AND METHODS: One hundred genetically well-characterized B-CLL samples, together with 11 healthy control samples, were analyzed using oligonucleotide arrays, which test for the expression of some 12,000 human genes. RESULTS: Aiming at microarray-based subclassification, class predictors were constructed using sets of differentially expressed genes, which yielded in zero or low misclassification rates. Furthermore, a significant number of the differentially expressed genes clustered in chromosomal regions affected by the respective genomic losses/gains. Deletions affecting chromosome bands 11q22-q23 and 17p13 led to a reduced expression of the corresponding genes, such as ATM and p53, while trisomy 12 resulted in the upregulation of genes mapping to chromosome arm 12q. Using an unsupervised analysis algorithm, expression profiling allowed partitioning into predominantly VH-mutated versus unmutated patient groups; however, association of the expression profile with the VH mutational status could only be detected in male patients. CONCLUSION: The finding that the most significantly differentially expressed genes are located in the corresponding aberrant chromosomal regions indicates that a gene dosage effect may exert a pathogenic role in B-CLL. The significant difference in the partitioning of male and female B-CLL samples suggests that the genomic signature for the VH mutational status might be sex-related.

Adult↗
PubMed2005

Histological heterogeneity of human glioblastomas investigated with an unsupervised neural network (SOM).

The histological variability of Glioblastomas (GB) precludes the modern assimilation of theses tumors into a single histological tumor group. As an alternative to statistical histological evaluation, we investigated 1489 human GB in order to discover whether they could be correctly classified using Self-Organizing Maps (SOM). In all tumors 50 histological features, as well as the age and sex of the patients, were examined. Four clusters of GB with a significance of 52 (maximal significance 60) were found. Cluster C1 contained 37.47% of all GB and 41.09% of all polymorphic glioblastomas (PG). Cluster C2 included 35.06% of all GB and 44.96% of all giant cell glioblastomas (GCG). Cluster C3 contained 16.45% of all GB with a significant component of astroblasts, glioblasts and oligodendroglia. Cluster C4 included 11.01% of all GB, 87.80% of the gliosarcomas (GS) and 36.72% of all GCG. Placing a series of component windows with their maps side by side allows the immediate recognition of the dependencies on variables and the determination of variables necessary to build the specific clusters. The SOM allow a realistic histological classification, comparable to the actual classification by the WHO. In addition, we found new, small subclusters of human GB which may have a clinical significance. With SOM one can learn to discriminate, discard and delete data, select histological and clinical or genetic variables that are meaningful, and consequently influence the result of patient management.

Adolescent↗
PubMed2006

Mining gene expression data by interpreting principal components.

BACKGROUND: There are many methods for analyzing microarray data that group together genes having similar patterns of expression over all conditions tested. However, in many instances the biologically important goal is to identify relatively small sets of genes that share coherent expression across only some conditions, rather than all or most conditions as required in traditional clustering; e.g. genes that are highly up-regulated and/or down-regulated similarly across only a subset of conditions. Equally important is the need to learn which conditions are the decisive ones in forming such gene sets of interest, and how they relate to diverse conditional covariates, such as disease diagnosis or prognosis. RESULTS: We present a method for automatically identifying such candidate sets of biologically relevant genes using a combination of principal components analysis and information theoretic metrics. To enable easy use of our methods, we have developed a data analysis package that facilitates visualization and subsequent data mining of the independent sources of significant variation present in gene microarray expression datasets (or in any other similarly structured high-dimensional dataset). We applied these tools to two public datasets, and highlight sets of genes most affected by specific subsets of conditions (e.g. tissues, treatments, samples, etc.). Statistically significant associations for highlighted gene sets were shown via global analysis for Gene Ontology term enrichment. Together with covariate associations, the tool provides a basis for building testable hypotheses about the biological or experimental causes of observed variation. CONCLUSION: We provide an unsupervised data mining technique for diverse microarray expression datasets that is distinct from major methods now in routine use. In test uses, this method, based on publicly available gene annotations, appears to identify numerous sets of biologically relevant genes. It has proven especially valuable in instances where there are many diverse conditions (10's to hundreds of different tissues or cell types), a situation in which many clustering and ordering algorithms become problematic. This approach also shows promise in other topic domains such as multi-spectral imaging datasets.

Algorithms↗
PubMed2003

Early unsupervised drinking--reducing the risks. The School Health and Alcohol Harm Reduction Project.

The School Health and Alcohol Harm Reduction Project (SHAHRP) aimed to reduce alcohol-related harm by enhancing students' abilities to identify and deal with high-risk drinking situations and issues. The SHAHRP study involved a quasi-experimental research design, incorporating intervention and control groups and measuring change over a 32-month period. The study occurred in metropolitan, government secondary schools (13 - 17-year-olds) in Perth, Western Australia. The 14 intervention and control schools involved in the SHAHRP study represent approximately 23% of government secondary schools in the Perth metropolitan area. The sample was selected using cluster sampling, with stratification by socio-economic area, and involved over 2,300 intervention and control students from junior secondary schools. The retention rate of the study was 75.9% over 32 months. The intervention incorporated evidence-based approaches to enhance potential for behaviour change in the target population. The intervention was a classroom-based programme, with an explicit harm minimization goal, and was conducted in two phases over a 2-year period. The results were analysed by baseline context of alcohol use to assess the impact of the programme on students with varying experience with alcohol. Knowledge and attitudes were modified simultaneously after the first phase of the intervention in all baseline context of use groups. The programme had little behavioural impact on baseline supervised drinkers; however, baseline non-drinkers and unsupervised drinkers were less likely to consume alcohol in a risky manner, compared to their corresponding control groups. In line with programme goals, early unsupervised drinkers from the intervention group were also significantly less likely to experience harm associated with their own use of alcohol compared to the corresponding control group. Unsupervised drinkers experienced 18.4% less alcohol-related harm after participating in both phases of the programme and this difference was maintained (19.4% difference) 17 months after the completion of the programme. This study indicates that a school drug education programme needs to be offered in several phases, that programme components may need to be included to cater for the differing baseline context of use groups, and that early unsupervised drinkers experience less alcohol-related harm after participating in a harm reduction programme.

Adolescent↗
PubMed2000

What harms do young Australians experience in alcohol-use situations?

OBJECTIVE: An insight into the alcohol-related experiences of young students in Perth, Western Australia, with particular emphasis to alcohol-related harm. METHOD: The sample of 2,329 students (female: n = 1,089, male: n = 1,240) is a school-based group selected using cluster sampling, with stratification by socio-economic area and represents 11 to 12 year olds' experiences with alcohol and alcohol-related harm. The SHAHRP survey instrument was developed and pre-tested to measure students' knowledge, attitudes, patterns and context of use, harms associated with the students' own alcohol consumption and harms associated with other people's use of alcohol and incorporates the students' perceptions of alcohol-related harm. RESULTS: Nearly two-thirds of all young people consumed alcohol under adult supervision; nearly 40% of all young males and 34% of all young females drink alcohol in unsupervised situations; and a fifth of young males consumed alcohol alone. Young males start drinking younger and consumed alcohol more regularly than young females, and consumed more alcohol per occasion. In the past 12 months, young males experienced more than five and young females more than three alcohol-related harms associated with their own alcohol consumption. They experienced a similar number of harms associated with other people's use of alcohol. Unsupervised drinkers were nearly seven times more likely to experience alcohol-related harm than supervised drinkers and nearly 13 times more likely to experience alcohol-related harm than non-drinkers. CONCLUSION AND IMPLICATIONS: The results can help inform the development of alcohol education programs for young people.

Age Distribution↗
PubMed2005

Delimitation of squamous cell cervical carcinoma using infrared microspectroscopic imaging.

Infrared (IR) spectroscopic imaging coupled with microscopy has been used to investigate thin sections of cervix uteri encompassing normal tissue, precancerous structures, and squamous cell carcinoma. Methods for unsupervised distinction of tissue types based on IR spectroscopy were developed. One-hundred and twenty-two images of cervical tissue were recorded by an FTIR spectrometer with a 64x64 focal plane array detector. The 499,712 IR spectra obtained were grouped by an approach which used fuzzy C-means clustering followed by hierarchical cluster analysis. The resulting false color maps were correlated with the morphological characteristics of an adjacent section of hematoxylin and eosin-stained tissue. In the first step, cervical stroma, epithelium, inflammation, blood vessels, and mucus could be distinguished in IR images by analysis of the spectral fingerprint region (950-1480 cm(-1)). In the second step, analysis in the spectral window 1420-1480 cm(-1) enables, for the first time, IR spectroscopic distinction between the basal layer, dysplastic lesions and squamous cell carcinoma within a particular sample. The joint application of IR microspectroscopic imaging and multivariate spectral processing combines diffraction-limited lateral optical resolution on the single cell level with highly specific and sensitive spectral classification on the molecular level. Compared with previous reports our approach constitutes a significant progress in the development of optical molecular spectroscopic techniques toward an additional diagnostic tool for the early histopathological characterization of cervical cancer.

Carcinoma, Squamous Cell↗
PubMed2005

Assessment of self-organizing maps to analyze sole-carbon source utilization profiles.

The use of community-level physiological profiles obtained with Biolog microplates is widely employed to consider the functional diversity of bacterial communities. Biolog produces a great amount of data which analysis has been the subject of many studies. In most cases, after some transformations, these data were investigated with classical multivariate analyses. Here we provided an alternative to this method, that is the use of an artificial intelligence technique, the Self-Organizing Maps (SOM, unsupervised neural network). We used data from a microcosm study of algae-associated bacterial communities placed in various nutritive conditions. Analyses were carried out on the net absorbances at two incubation times for each substrates and on the chemical guild categorization of the total bacterial activity. Compared to Principal Components Analysis and cluster analysis, SOM appeared as a valuable tool for community classification, and to establish clear relationships between clusters of bacterial communities and sole-carbon sources utilization. Specifically, SOM offered a clear bidimensional projection of a relatively large volume of data and were easier to interpret than plots commonly obtained with multivariate analyses. They would be recommended to pattern the temporal evolution of communities' functional diversity.

Bacteria↗
PubMed2025

Tumor-immune partitioning and clustering algorithm for identifying tumor-immune cell spatial interaction signatures within the tumor microenvironment.

BACKGROUND: Growing evidence supports the importance of characterizing the organizational patterns of various cellular constituents in the tumor microenvironment in precision oncology. Most existing data on immune cell infiltrates in tumors, which are based on immune cell counts or nearest neighbor-type analyses, have failed to fully capture the cellular organization and heterogeneity. METHODS: We introduce a computational algorithm, termed Tumor-Immune Partitioning and Clustering (TIPC), that jointly measures immune cell partitioning between tumor epithelial and stromal areas and immune cell clustering versus dispersion. As proof-of-principle, we applied TIPC to a prospective cohort incident tumor biobank containing 931 colorectal carcinoma cases. TIPC identified tumor subtypes with unique spatial patterns between tumor cells and T lymphocytes linked to certain molecular pathologic and prognostic features. T lymphocyte identification and phenotyping were achieved using multiplexed (multispectral) immunofluorescence. In a separate hepatocellular carcinoma cohort, we replaced the stromal component with specific immune cell types-CXCR3+CD68+ or CD8+-to profile their spatial relationships with CXCL9+CD68+ cells. RESULTS: Six unsupervised TIPC subtypes based on T lymphocyte distribution patterns were identified, comprising two cold and four hot subtypes. Three of the four hot subtypes were associated with significantly longer colorectal cancer (CRC)-specific survival compared to a reference cold subtype. Our analysis showed that variations in T-cell densities among the TIPC subtypes did not strictly correlate with prognostic benefits, underscoring the prognostic significance of immune cell spatial patterns. Additionally, TIPC revealed two spatially distinct and cell density-specific subtypes among microsatellite instability-high colorectal cancers, indicating its potential to upgrade tumor subtyping. TIPC was also applied to additional immune cell types, eosinophils and neutrophils, identified using morphology and supervised machine learning; here two tumor subtypes with similarly low densities, namely 'cold, tumor-rich' and 'cold, stroma-rich', exhibited differential prognostic associations. Lastly, we validated our methods and results using The Cancer Genome Atlas colon and rectal adenocarcinoma data (n = 570). Moreover, applying TIPC to hepatocellular carcinoma cases (n = 27) highlighted critical cell interactions like CXCL9-CXCR3 and CXCL9-CD8. CONCLUSIONS: Unsupervised discoveries of microgeometric tissue organizational patterns and novel tumor subtypes using the TIPC algorithm can deepen our understanding of the tumor immune microenvironment and likely inform precision cancer immunotherapy.

Humans↗
PubMed2004

Improved fourier transform method for unsupervised cell-cycle regulated gene prediction.

MOTIVATION: Cell-cycle regulated gene prediction using microarray time-course measurements of the mRNA expression levels of genes has been used by several researchers. The popularly employed approach is Fourier transform (FT) method in conjunction with the set of known cell-cycle regulated genes. In the absence of training data, fourier transform method is sensitive to noise, additive monotonic component arising from cell population growth and deviation from strict sinusoidal form of expression. Known cell cycle regulated genes may not be available for certain organisms or using them for training may bias the prediction. RESULTS: In this paper we propose an Improved Fourier Transform (IFT) method which takes care of several factors such as monotonic additive component of the cell-cycle expression, irregular or partial-cycle sampling of gene expression. The proposed algorithm does not need any known cell-cycle regulated genes for prediction. Apart from alleviating need for training set, it also removes bias towards genes similar to the training set. We have evaluated the developed method on two publicly available datasets: yeast cell-cycle data and HeLa cell-cycle data. The proposed algorithm has performed competitively on both datasets with that of the supervised fourier transform method used. It outperformed other unsupervised methods such as Partial Least Squares (PLS) and Single Pulse Modeling (SPM). This method is easy to comprehend and implement, and runs faster.

Algorithms↗
PubMed2003

Venn Mapping: clustering of heterologous microarray data based on the number of co-occurring differentially expressed genes.

MOTIVATION: To evaluate microarray data, clustering is widely used to group biological samples or genes. However, problems arise when comparing heterologous databases. As the clustering algorithm searches for similarities between experiments, it will most likely first separate the data sets, masking relationships that exist between samples from different databases. RESULTS: We developed a program, Venn Mapper, to calculate the statistical significance of the number of co-occurring differentially expressed genes in any of the two experiments. For proof of principle, we analysed a heterologous data set of 170 microarrays including breast and prostate cancer microarray analyses. Significant overlap was found in an unsupervised analysis between metastasized prostate cancer and metastasized breast cancer and BRCA mutated breast cancer. A comparison between single microarray data and the averaged breast and prostate data sets was also evaluated. This analysis suggests that genes expressed higher in stromal cells are also implicated in metastatic prostate cancer and BRCA mutated breast cancer. The Venn Mapper program identifies overlaps between samples from heterologous data sets and directly extracts the genes responsible for the overlap. From this information novel biological hypotheses may be addressed. AVAILABILITY: Venn Mapper is freely available on http://www.erasmusmc.nl/gatcplatform. SUPPLEMENTARY INFORMATION: http://www.erasmusmc.nl/gatcplatform/vennmapper.html.

Algorithms↗
PubMedDate not supplied

Transferring of speech movements from video to 3D face space.

We present a novel method for transferring speech animation recorded in low quality videos to high resolution 3D face models. The basic idea is to synthesize the animated faces by an interpolation based on a small set of 3D key face shapes which span a 3D face space. The 3D key shapes are extracted by an unsupervised learning process in 2D video space to form a set of 2D visemes which are then mapped to the 3D face space. The learning process consists of two main phases: 1) Isomap-based nonlinear dimensionality reduction to embed the video speech movements into a low-dimensional manifold and 2) K-means clustering in the low-dimensional space to extract 2D key viseme frames. Our main contribution is that we use the Isomap-based learning method to extract intrinsic geometry of the speech video space and thus to make it possible to define the 3D key viseme shapes. To do so, we need only to capture a limited number of 3D key face models by using a general 3D scanner. Moreover, we also develop a skull movement recovery method based on simple anatomical structures to enhance 3D realism in local mouth movements. Experimental results show that our method can achieve realistic 3D animation effects with a small number of 3D key face models.

Algorithms↗
PubMed2001

Identifying splits with clear separation: a new class discovery method for gene expression data.

We present a new class discovery method for microarray gene expression data. Based on a collection of gene expression profiles from different tissue samples, the method searches for binary class distinctions in the set of samples that show clear separation in the expression levels of specific subsets of genes. Several mutually independent class distinctions may be found, which is difficult to obtain from most commonly used clustering algorithms. Each class distinction can be biologically interpreted in terms of its supporting genes. The mathematical characterization of the favored class distinctions is based on statistical concepts. By analyzing three data sets from cancer gene expression studies, we demonstrate that our method is able to detect biologically relevant structures, for example cancer subtypes, in an unsupervised fashion.

Algorithms↗