Subclass mapping: Identifying common subtypes in independent disease data sets

Yujin Hoshida, Jean Philippe Brunet, Pablo Tamayo, Todd R. Golub, Jill P. Mesirov

Research output: Contribution to journalArticle

101 Citations (Scopus)

Abstract

Whole genome expression profiles are widely used to discover molecular subtypes of diseases. A remaining challenge is to identify the correspondence or commonality of subtypes found in multiple, independent data sets generated on various platforms. While model-based supervised learning is often used to make these connections, the models can be biased to the training data set and thus miss inherent, relevant substructure in the test data. Here we describe an unsupervised subclass mapping method (SubMap), which reveals common subtypes between independent data sets. The subtypes within a data set can be determined by unsupervised clustering or given by predetermined phenotypes before applying SubMap. We define a measure of correspondence for subtypes and evaluate its significance building an our previous work on gene set enrichment analysis. The strength of the SubMap method is that it does not impose the structure of one data set upon another, but rather uses a bi-directional approach to highlight the common substructures in both. We show how this method can reveal the correspondence between several cancer-related data sets. Notably, it identifies common subtypes of breast cancer associated with estrogen receptor status, and a subgroup of lymphoma patients who share similar survival patterns, thus improving the accuracy of a clinical outcome predictor.

Original languageEnglish (US)
Article numbere1195
JournalPloS one
Volume2
Issue number11
DOIs
StatePublished - Nov 21 2007
Externally publishedYes

Fingerprint

Genes
Supervised learning
Estrogen Receptors
methodology
lymphoma
breast neoplasms
learning
Cluster Analysis
Datasets
Lymphoma
phenotype
neoplasms
Learning
Genome
genome
Breast Neoplasms
Phenotype
Survival
genes
testing

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)

Cite this

Subclass mapping : Identifying common subtypes in independent disease data sets. / Hoshida, Yujin; Brunet, Jean Philippe; Tamayo, Pablo; Golub, Todd R.; Mesirov, Jill P.

In: PloS one, Vol. 2, No. 11, e1195, 21.11.2007.

Research output: Contribution to journalArticle

Hoshida, Yujin ; Brunet, Jean Philippe ; Tamayo, Pablo ; Golub, Todd R. ; Mesirov, Jill P. / Subclass mapping : Identifying common subtypes in independent disease data sets. In: PloS one. 2007 ; Vol. 2, No. 11.
@article{357c27ee3d314b92be687f24ba7a5a71,
title = "Subclass mapping: Identifying common subtypes in independent disease data sets",
abstract = "Whole genome expression profiles are widely used to discover molecular subtypes of diseases. A remaining challenge is to identify the correspondence or commonality of subtypes found in multiple, independent data sets generated on various platforms. While model-based supervised learning is often used to make these connections, the models can be biased to the training data set and thus miss inherent, relevant substructure in the test data. Here we describe an unsupervised subclass mapping method (SubMap), which reveals common subtypes between independent data sets. The subtypes within a data set can be determined by unsupervised clustering or given by predetermined phenotypes before applying SubMap. We define a measure of correspondence for subtypes and evaluate its significance building an our previous work on gene set enrichment analysis. The strength of the SubMap method is that it does not impose the structure of one data set upon another, but rather uses a bi-directional approach to highlight the common substructures in both. We show how this method can reveal the correspondence between several cancer-related data sets. Notably, it identifies common subtypes of breast cancer associated with estrogen receptor status, and a subgroup of lymphoma patients who share similar survival patterns, thus improving the accuracy of a clinical outcome predictor.",
author = "Yujin Hoshida and Brunet, {Jean Philippe} and Pablo Tamayo and Golub, {Todd R.} and Mesirov, {Jill P.}",
year = "2007",
month = "11",
day = "21",
doi = "10.1371/journal.pone.0001195",
language = "English (US)",
volume = "2",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "11",

}

TY - JOUR

T1 - Subclass mapping

T2 - Identifying common subtypes in independent disease data sets

AU - Hoshida, Yujin

AU - Brunet, Jean Philippe

AU - Tamayo, Pablo

AU - Golub, Todd R.

AU - Mesirov, Jill P.

PY - 2007/11/21

Y1 - 2007/11/21

N2 - Whole genome expression profiles are widely used to discover molecular subtypes of diseases. A remaining challenge is to identify the correspondence or commonality of subtypes found in multiple, independent data sets generated on various platforms. While model-based supervised learning is often used to make these connections, the models can be biased to the training data set and thus miss inherent, relevant substructure in the test data. Here we describe an unsupervised subclass mapping method (SubMap), which reveals common subtypes between independent data sets. The subtypes within a data set can be determined by unsupervised clustering or given by predetermined phenotypes before applying SubMap. We define a measure of correspondence for subtypes and evaluate its significance building an our previous work on gene set enrichment analysis. The strength of the SubMap method is that it does not impose the structure of one data set upon another, but rather uses a bi-directional approach to highlight the common substructures in both. We show how this method can reveal the correspondence between several cancer-related data sets. Notably, it identifies common subtypes of breast cancer associated with estrogen receptor status, and a subgroup of lymphoma patients who share similar survival patterns, thus improving the accuracy of a clinical outcome predictor.

AB - Whole genome expression profiles are widely used to discover molecular subtypes of diseases. A remaining challenge is to identify the correspondence or commonality of subtypes found in multiple, independent data sets generated on various platforms. While model-based supervised learning is often used to make these connections, the models can be biased to the training data set and thus miss inherent, relevant substructure in the test data. Here we describe an unsupervised subclass mapping method (SubMap), which reveals common subtypes between independent data sets. The subtypes within a data set can be determined by unsupervised clustering or given by predetermined phenotypes before applying SubMap. We define a measure of correspondence for subtypes and evaluate its significance building an our previous work on gene set enrichment analysis. The strength of the SubMap method is that it does not impose the structure of one data set upon another, but rather uses a bi-directional approach to highlight the common substructures in both. We show how this method can reveal the correspondence between several cancer-related data sets. Notably, it identifies common subtypes of breast cancer associated with estrogen receptor status, and a subgroup of lymphoma patients who share similar survival patterns, thus improving the accuracy of a clinical outcome predictor.

UR - http://www.scopus.com/inward/record.url?scp=43149108659&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=43149108659&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0001195

DO - 10.1371/journal.pone.0001195

M3 - Article

C2 - 18030330

AN - SCOPUS:43149108659

VL - 2

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 11

M1 - e1195

ER -