Bayesian consensus clustering

Lock, Eric F.; Dunson, David B.

doi:10.1093/bioinformatics/btt425

Cited by 237 publications

(226 citation statements)

References 24 publications

Supporting

Mentioning

217

Contrasting

Order By: Relevance

“…Spatial analysis was performed in the AL cases dataset to find clusters according to their location using the Density-Based Spatial Clustering of Applications with Noise (DB-SCAN) [16][17][18] algorithm on R statistical software 19,20 using the flexible procedures for clustering package 21 . This algorithm identifies clusters by finding point subsets, which have a minimum number of points that are density-connected from each point's neighborhood within a given radius.…”

Section: Spatial Data Analysismentioning

confidence: 99%

Geographical Distribution and Cluster Detection of Childhood Leukemia in the Metropolitan Area of Guadalajara, Mexico

Tlacuilo‐Parra

Garibaldi-Covarrubias

Romo-Rubio

et al. 2017

RIC

View full text Add to dashboard Cite

Background: Acute leukemia is the most common cancer in childhood. Analyzing the spatial distribution of acute leukemia may generate the identification of risk factors. Objective: To study the incidence rate of acute leukemia, its geographic distribution, and cluster detection in the metropolitan area of Guadalajara, Mexico. Methods: We included children under 15 years of age diagnosed with acute leukemia during the period 2010-2014 in the metropolitan area of Guadalajara. Each case was geo-referenced to street level to latitude and longitude coordinates using Quantum Geographic Information System (QGIS). Spatial clusters were found in the location of the acute leukemia cases applying the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm with R statistical software. Results: A total of 269 cases of leukemia were registered, 227 (84%) were acute lymphoblastic leukemia and 42 (16%) acute myeloblastic leukemia. The mean age was 6 ± 4 years. The mean incidence of acute leukemia was 6.44 cases/100,000 inhabitants: El Salto 10.12/100,000, Guadalajara 7.55/100,000, and Tlaquepaque 6.74/100,000. The DBSCAN found three clusters, all located within the municipality of Guadalajara. Conclusions: The incidence of acute leukemia in our population is higher than that in Canada and the USA. We found three spatial clusters of childhood acute lymphoblastic leukemia in the municipality of Guadalajara, suggesting the presence of local predisposing factors. (REV INVES CLIN. 2017;69:159-65)

show abstract

Section: Spatial Data Analysismentioning

confidence: 99%

Geographical Distribution and Cluster Detection of Childhood Leukemia in the Metropolitan Area of Guadalajara, Mexico

Tlacuilo‐Parra

Garibaldi-Covarrubias

Romo-Rubio

et al. 2017

RIC

View full text Add to dashboard Cite

show abstract

“…Out of the 348 samples, there are 5 subtypes of breast cancer with different number of samples in each subtype: Basal (66), Her2 (42), LumA (154), LumB (81), and Normal (5). We preprocessed the data in the same way as in [21]. We first imputed missing values with the k-nearest neighbors algorithm (k = 10), then removed genes with low variations across samples (standard deviation smaller than 1.5), and finally mean centered each gene.…”

Section: Breast Cancer Datamentioning

confidence: 99%

Supervised singular value decomposition and its asymptotic properties

Yang

Nobel

et al. 2016

Journal of Multivariate Analysis

View full text Add to dashboard Cite

a b s t r a c tA supervised singular value decomposition (SupSVD) model has been developed for supervised dimension reduction where the low rank structure of the data of interest is potentially driven by additional variables measured on the same set of samples. The SupSVD model can make use of the information in the additional variables to accurately extract underlying structures that are more interpretable. The model is general and includes the principal component analysis model and the reduced rank regression model as two extreme cases. The model is formulated in a hierarchical fashion using latent variables, and a modified expectation-maximization algorithm for parameter estimation is developed, which is computationally efficient. The asymptotic properties for the estimated parameters are derived. We use comprehensive simulations and a real data example to illustrate the advantages of the SupSVD model.

show abstract

“…While integration of multiple omics data sources from the same cohort provides meaningful insight into the molecular and cellular processes of the disease and hence becomes popular, the problem brings new challenges in data analysis; yet the development of statistical methods in this field has just started. Similar to traditional microarray data analysis, vertical omics integration can target on the following biological purposes: (i) candidate marker detection (Wang and others, 2012); (ii) gene set or pathway analysis (Hu and Tzeng, 2014); (iii) dimension reduction (Lock and others, 2013); (iv) classification analysis (Seoane and others, 2014); and finally (v) clustering analysis to identify disease subtypes others, 2009, 2013;Lock and Dunson, 2013). The last objective (clustering analysis) is the focus of this paper.…”

Section: Introductionmentioning

confidence: 99%

“…Rey and Roth (2012) introduced a copula mixture model for dependency-seeking clustering of multi-omics data. Lock and Dunson (2013) proposed a Bayesian consensus clustering to account for consensus and source-specific information when identifying clusters. others (2009, 2013) developed an integrative clustering (iCluster) approach via Gaussian latent regression model.…”

Section: Introductionmentioning

confidence: 99%

Integrative clustering of multi-level omics data for disease subtype discovery using sequential double regularization

Kim

Oesterreich

Kim

et al. 2016

Biostat

View full text Add to dashboard Cite

SUMMARYWith the rapid advances in technologies of microarray and massively parallel sequencing, data of multiple omics sources from a large patient cohort are now frequently seen in many consortium studies. Effective multi-level omics data integration has brought new statistical challenges. One important biological objective of such integrative analysis is to cluster patients in order to identify clinically relevant disease subtypes, which will form basis for tailored treatment and personalized medicine. Several methods have been proposed in the literature for this purpose, including the popular iCluster method used in many cancer applications. When clustering high-dimensional omics data, effective feature selection is critical for better clustering accuracy and biological interpretation. It is also common that a portion of "scattered samples" has patterns distinct from all major clusters and should not be assigned into any cluster as they may represent a rare disease subcategory or be in transition between disease subtypes. In this paper, we firstly propose to improve feature selection of the iCluster factor model by an overlapping sparse group lasso penalty on the omics features using prior knowledge of inter-omics regulatory flows. We then perform regularization over samples to allow clustering with scattered samples and generate tight clusters. The proposed group structured tight iCluster method will be evaluated by two real breast cancer examples and simulations to demonstrate its improved clustering accuracy, biological interpretation, and ability to generate coherent tight clusters.

show abstract

Bayesian consensus clustering

Abstract: R code with instructions and examples is available at http://people.duke.edu/%7Eel113/software.html.

Cited by 237 publications

References 24 publications

Geographical Distribution and Cluster Detection of Childhood Leukemia in the Metropolitan Area of Guadalajara, Mexico

Geographical Distribution and Cluster Detection of Childhood Leukemia in the Metropolitan Area of Guadalajara, Mexico

Supervised singular value decomposition and its asymptotic properties

Integrative clustering of multi-level omics data for disease subtype discovery using sequential double regularization

Contact Info

Product

Resources

About