glioma: Array CGH data of a glioblastoma tumor

In pbreheny/hdrm: High-Dimensional Regression Modeling

Array CGH data of a glioblastoma tumor

Description

Broadly speaking, humans have two copies of their genome. Occasionally however, a region of the genome is duplicated or destroyed; this is known as copy number variation. All humans have some degree of copy number variation; many cancers, however, display much more extreme variation. Comparative genomic hybridization (CGH) is one method for measuring the number of copies present at a genome-wide scale.

This data set consists of CGH data from glioblastoma tumors. In the data set, the data from two tumors (chromosome 7 in one patient, chromosome 13 in another) are spliced together in order to create a challenging data for CNV detection consisting of both gains and losses over both short and large scales. It has become a popular benchmark data set for CNV studies.

Format

• The numbers in the vector are the log (base 2) ratio of the number of DNA copies of the tumor relative to a normal reference. Values above 0 represent copy number gains, while values less than 0 represent copy number losses.

Dimensions

• 990 features

• 1 instance; note that since there is only 1 instance, the data is simply stored a numeric vector.

References

The original data appear in:

Bredel M, Bredel C, Juric D, Harsh GR, Vogel H, Recht LD, and Sikic BI. (2005). High-Resolution Genome-Wide Mapping of Genetic Alterations in Human Glial Brain Tumors. Cancer Research, 65: 4088-4096.

The pseudo-chromosome data set was introduced by:

Tibshirani R, and Wang P. (2008). Spatial smoothing and hot spot detection for CGH data using the fused lasso. Biostatistics, 9: 18-29.

The data used to be available in the cghFLasso package, but this package appears to be no longer available.

pbreheny/hdrm documentation built on Jan. 17, 2024, 8:53 p.m.