Dataset Information

Supervised harvesting of expression trees.

ABSTRACT: BACKGROUND:We propose a new method for supervised learning from gene expression data. We call it 'tree harvesting'. This technique starts with a hierarchical clustering of genes, then models the outcome variable as a sum of the average expression profiles of chosen clusters and their products. It can be applied to many different kinds of outcome measures such as censored survival times, or a response falling in two or more classes (for example, cancer classes). The method can discover genes that have strong effects on their own, and genes that interact with other genes. RESULTS:We illustrate the method on data from a lymphoma study, and on a dataset containing samples from eight different cancers. It identified some potentially interesting gene clusters. In simulation studies we found that the procedure may require a large number of experimental samples to successfully discover interactions. CONCLUSIONS:Tree harvesting is a potentially useful tool for exploration of gene expression data and identification of interesting clusters of genes worthy of further investigation.

SUBMITTER: Hastie T

PROVIDER: S-EPMC17599 | biostudies-literature | 2001

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Supervised harvesting of expression trees.

Hastie T T Tibshirani R R Botstein D D Brown P P

Genome biology 20010110 1

<h4>Background</h4>We propose a new method for supervised learning from gene expression data. We call it 'tree harvesting'. This technique starts with a hierarchical clustering of genes, then models the outcome variable as a sum of the average expression profiles of chosen clusters and their products. It can be applied to many different kinds of outcome measures such as censored survival times, or a response falling in two or more classes (for example, cancer classes). The method can discover ge ...[more]

PMID: 11178280

Similar Datasets

Project description:BACKGROUND: Gene expression studies can be used to help identify disease-associated genes by comparing the levels of expressed transcripts between cases and controls, and to identify functional genetic variants (expression quantitative loci or eQTLs) by comparing expression levels between individuals with different genotypes. While many of these studies are performed in blood or lymphoblastoid cell lines due to tissue accessibility, the relevance of expression differences in tissues that are not the primary site of disease is unclear. Further, many eQTLs are tissue specific. Thus, there is a clear and compelling need to conduct gene expression studies in tissues that are specifically relevant to the disease of interest. One major technical concern about using autopsy-derived tissue is how representative it is of physiologic conditions, given the effect of postmortem interval on tissue degradation. RESULTS: In this study, we monitored the gene expression of 13 tissue samples harvested from a rapid autopsy heart (non-failed heart) and 7 from a cardiac explant (failed heart) through 24 hours of autolysis. The 24 hour autopsy simulation was designed to reflect a typical autopsy scenario where a body may begin cooling to ambient temperature for ~12 hours, before transportation and storage in a refrigerated room in a morgue. In addition, we also simulated a scenario wherein the body was left at room temperature for up to 24 hours before being found. A small fraction (< 2.5%) of genes showed fluctuations in expression over the 24 hr period and largely belong to immune and signal response and energy metabolism-related processes. Global expression analysis suggests that RNA expression is reproducible over 24 hours of autolysis with 95% genes showing < 1.2 fold change. Comparing the rapid autopsy to the failed heart identified 480 differentially expressed genes, including several types of collagens, lumican (LUM), natriuretic peptide A (NPPA) and connective tissue growth factor (CTGF), which allows for the clear separation between failing and non-failing heart based on gene expression profiles. CONCLUSIONS: Our results demonstrate that RNA from autopsy-derived tissue, even up to 24 hours of autolysis, can be used to identify biologically relevant expression pattern differences, thus serving as a practical source for gene expression experiments.

Dataset Information

Supervised harvesting of expression trees.

Publications

Supervised harvesting of expression trees.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets