leukemia {supclust}

A part of the Golub's famous AML/ALL-leukemia dataset

Description

Part of the training set of the famous AML/ALL-leukemia dataset from the Whitehead Institute. It has been reduced to 250 genes, about half of which are very informative for classification, whereas the other half was chosen randomly.

Usage

data(leukemia)

Format

Contains three R-objects: The expression (38 x 250) matrix leukemia.x, the associated binary (0,1) response variable leukemia.y, and the associated 3-class response variable leukemia.z with values in 0,1,2.

Author(s)

Marcel Dettling

Source

originally at http://www.genome.wi.mit.edu/MPR/, (which is not a valid URL any more).

References

First published in
Golub et al. (1999) Molecular Classification of Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring. Science 286, 531–538.

Examples

data(leukemia, package="supclust")
str(leukemia.x)
str(leukemia.y)
str(leukemia.z)
op <- par(mfrow= 1:2)
plot(leukemia.x[,56], leukemia.y)
plot(leukemia.x[,174],leukemia.z)
par(op)