Unsupervised Selection of Highly Coexpressed and Noncoexpressed Genes Using a Consensus Clustering Approach

Abstract

In this paper we explore the concept of consensus clustering to identify, within a set of differentially expressed genes, a subset of genes that are either highly coexpressed or highly noncoexpressed based on the hypothesis that this subset would serve as a better starting point for further analyses. A number of core clustering methods form the basis for the assertion of an agreement matrix (AM) characterizing the level of coexpression between any two probesets. In order to overcome the limitations of using a single distance metric, we explore different metrics and examine the sensitivity of the AM as a function of the input number of clusters to find a suggestive number of clusters that best describes a particular dataset. The result of this level of analysis is a systematic framework for eliminating probesets that cannot be clearly characterized as either coexpressed or noncoexpressed with others, thus eliminating a number of probesets from further analysis. Subsequently, an agglomerative hierarchical clustering approach is applied to cluster the selected subset using the agreement metric information as the similarity measure. Thus, the goal of the proposed methodology is twofold: (1) we opt to identify a more "clusterable" subset of the original set; and (2) we aim at further refining the subset in order to identify a core of genes that contains genes that are either coexpressed or noncoexpressed within a certain confidence level. The approach is tested with a number of data sets, both synthetic and real, and it is demonstrated that it is successful in identifying more clusterable, also hypothesized to be more biologically relevant, subsets of genes and expression profiles.

Dynamics of Short-Term Gene Expression Profiling in Liver Following Thermal Injury

“…This step determined a set of probesets whose expression patterns were significantly altered following the treatment considering the temporal differences between the control and treatment groups. Finally the data sets corresponding to those differentially expressed probesets in either burn and/ or sham groups were combined to form one single matrix, which was then clustered using the a approach “consensus clustering” (21), in an unsupervised manner. This provided a set of burn responsive genes, which is significantly different than that of control group.…”

Tandem Analysis of Transcriptome and Proteome Changes after a Single Dose of Corticosteroid: A Systems Approach to Liver Function in Pharmacogenomics

“…While hierarchical clustering analysis described above identifies the potential co-regulatory schemes for the genes in the intersection of transcriptomic and proteomic datasets; it fails to capture the dynamics in the rest of the genes that may also show differences in expression over time, although they may not co-exist in both datasets. In order to evaluate the overall dynamic patterns and extract the most useful information integrating these two datasets, a consensus clustering (Nguyen et al, 2009) method was applied to these two ''-omic'' datasets separately. First, proteins with differential temporal profiles were clustered using p values of 0.05 for significant clusters and an agreement level of 0.70 for the genes in each cluster.…”

Circadian Signatures in Rat Liver: From Gene Expression to Pathways

“…To study the effect of individual gene expression on the pathway activity level, we depict the relationship between the weights and the correlation of the individual genes (the correlation between gene expression levels and Figure 3 The five significant clusters identified by a consensus clustering analysis [19] using δ = 0.65. The pathway activity level (PAL) of pathways represents the presented curves and the exact reverse curves; PAL = (-) PAL.…”