Cluster stability scores for microarray data in cancer studies

Abstract

BackgroundA potential benefit of profiling of tissue samples using microarrays is the generation of molecular fingerprints that will define subtypes of disease. Hierarchical clustering has been the primary analytical tool used to define disease subtypes from microarray experiments in cancer settings. Assessing cluster reliability poses a major complication in analyzing output from clustering procedures. While most work has focused on estimating the number of clusters in a dataset, the question of stability of individual-level clusters has not been addressed.ResultsWe address this problem by developing cluster stability scores using subsampling techniques. These scores exploit the redundancy in biologically discriminatory information on the chip. Our approach is generic and can be used with any clustering method. We propose procedures for calculating cluster stability scores for situations involving both known and unknown numbers of clusters. We also develop cluster-size adjusted stability scores. The method is illustrated by application to data three cancer studies; one involving childhood cancers, the second involving B-cell lymphoma, and the final is from a malignant melanoma study.AvailabilityCode implementing the proposed analytic method can be obtained at the second author's website.

A Modular Organization of the Human Intestinal Mucosal Microbiota and Its Association with Inflammatory Bowel Disease

“…To quantitatively evaluate the degree of module preservation, we carried out a Z-summary test (Methods). Alternative statistics are available to assess the quality and reproducibility of clusters among datasets [ 33 , 54 - 58 ]. An advantage of the Zsummary statistic is that it allows for significance thresholds: Z-summary <2 indicates no significant module preservation; 2<Z-summary<10 indicates moderate preservation; and, Z-summary>10 indicates strong preservation.…”

Microgeographic Proteomic Networks of the Human Colonic Mucosa and Their Association With Inflammatory Bowel Disease

“…Alternative statistics are available to assess the quality and reproducibility of clusters among data sets. 30 , 31 , 32 , 33 Our previous work outlined 7 simulation scenarios in which we compared WGCNA's module preservation statistics with the hitherto best-performing alternative approach, and the Z-summary statistic had distinct advantages when it came to studying the preservation of network modules. 29 , 34 Detailed discussions of the pros and cons of these statistics were presented recently.…”

Predicting Survival within the Lung Cancer Histopathological Hierarchy Using a Multi-Scale Genomic Model of Development

“…However, even though the molecular profiles of samples from one cancer subtype tend to occupy a common dendrogram cluster, there is significant admixing with samples from different cancer subtypes. Cluster configurations resulting from hierarchical algorithms are generally not well defined since they vary depending upon the input order of the same data [ 15 , 16 ] and are often irreproducible [ 17 ]. More importantly, hierarchical clustering (or any purely correlative technique) alone does not provide a rational biological basis for disease classification.…”