Untitled Document

PAM: Prediction Analysis for Microarrays
Class Prediction and Survival Analysis for Genomic Expression Data Mining

Features:

Performs sample classification from gene expression data,
via "nearest shrunken centroid method'' of Tibshirani, Hastie, Narasimhan and Chu (2002):
"Diagnosis of multiple cancer types by shrunken centroids of gene expression" (PNAS website).
PNAS 2002 99:6567-6572 (May 14).

For survival outcomes, implements 'supervised principal components' method. See
Semi-supervised methods for predicting patient survival from gene expression papers (Bair and Tibshirani) PLOS Biology, and Prediction by supervised principal components (Bair, Hastie, Paul, Tibshirani) Stanford tech report

Version 2.1 (Sep 14, 2005) featuring False discovery rates FDRs

Version 2.0 (Mar 7, 2005) featuring: FDRs and survival analysis via supervised principal components,

Estimates prediction error via cross-validation

Provides a list of significant genes whose expression characterizes each diagnostic class

Works with data from both cDNA and oligo microarrays. Can also be applied to protein expression data and SNP chip data.

What is nearest shrunken centroids?
How does it compare to other classifiers?

Developed at Stanford University Labs. Free for all users.

Yahoo newsgroup

Two versions:

Excel Add-in: Registration page; Installation guide; Manual;
PAM for the R package Superpc for the R package

Related links:

SAM package for assessing gene significance in microarrays
CGH-Miner package for CGH data;
PPC package for protein mass spec classification