Diagnosis of multiple cancer types by shrunken centroids of gene expression

doi:10.1073/pnas.082099299

Diagnosis of multiple cancer types by shrunken centroids of gene expression

Departments of ^†Health, Research and Policy, and Statistics, ^§Statistics and Health, Research and Policy, and ^¶Medicine and Biochemistry, Stanford University, Stanford, CA 94305

Communicated by Bradley Efron, Stanford University, Stanford, CA (received for review October 10, 2001)

Abstract

We have devised an approach to cancer class prediction from gene expression profiling, based on an enhancement of the simple nearest prototype (centroid) classifier. We shrink the prototypes and hence obtain a classifier that is often more accurate than competing methods. Our method of “nearest shrunken centroids” identifies subsets of genes that best characterize each class. The technique is general and can be used in many other classification problems. To demonstrate its effectiveness, we show that the method was highly efficient in finding genes for classifying small round blue cell tumors and leukemias.

Footnotes

↵ ‡ To whom reprint requests should be addressed. E-mail: tibs{at}stat.stanford.edu.
Abbreviations:

SRBCT,

small round blue cell tumors;

BL,

Burkitt lymphoma;

EWS,

Ewing sarcoma;

NB,

neuroblastoma;

RMS,

rhabdomyosarcoma