Bayesian infinite mixture model based clustering of gene expression profiles

Abstract

medvedm@email.uc.edu

Microarray analysis of cytoplasmic versus whole cell RNA reveals a considerable number of missed and false positive mRNAs

“…Cluster analysis was performed using the Bayesian infinite mixture models (Medvedovic and Sivaganesan 2002) as implemented in the GIMM software (http://eh3.uc.edu/gimm). Significantly differentially expressed genes were annotated with functional assignments to help determine which gene categories were enriched with differentially expressed genes using EASE, and the gene categories used were GenMAPP and the three branches of the and Gene Ontology (GO) database: biological process, molecular function, and cellular component (Ashburner et al 2000).…”

Multivariate Models and Algorithms for Learning Correlation Structures from Replicated Molecular Profiling Data

“…As the MLE's of parameters have closed-form representation, the model is computationally very efficient, e.g. it is well known that the infinite Bayesian mixture model approach (Medvedovic & Sivaganesan, 2002;Medvedovic et al, 2004) suffers from non-trivial computational complexity as the number of genes and replicated measurements increases. However, blind-case model always imposes a fixed number of parameters in the model.…”

“…Consequently, there has been an increasing interest in the development of computational methods to uncover gene association patterns underlying such data, e.g. gene clustering (Medvedovic & Sivaganesan, 2002;Medvedovic et al, 2004), inference of gene association networks (Altay and Emmert-Streib, 2010;Butte & Kohane, 2000;Zhu et al, 2005), sample classification (Yeung & Bumgarner, 2005) and detection of differentially expressed genes (Sartor et al, 2006). However, outcome of any bioinformatics analysis is directly influenced by the quality of molecular profiling data, which are often contaminated with excessive noise.…”

A Generalized Multivariate Approach to Pattern Discovery from Replicated and Incomplete Genome-Wide Measurements

“…With few exceptions, e.g., [21], [22], [40], the existing gene clustering and networking algorithms do not appropriately accommodate replicated and/or incomplete measurements. The increased power of detecting hidden patterns in data is achieved by sufficiently exploiting the replicates, which has been demonstrated using Infinite Bayesian Mixture Models (IBMM) [21], [22] and Parsimonious Multivariate Gaussian Models (PMGM) [40].…”

“…The increased power of detecting hidden patterns in data is achieved by sufficiently exploiting the replicates, which has been demonstrated using Infinite Bayesian Mixture Models (IBMM) [21], [22] and Parsimonious Multivariate Gaussian Models (PMGM) [40]. In the IBMM approaches, authors used both an "elliptical" model that allows within-replicate variation across difference samples to be different and a "spherical" model that does not.…”