PAM is a technique for sample classification from microarray data. This package features an Excel interface, designed and written by Balasubramanian Narasimhan, with help from Trevor Hastie and Rob Tibshirani.
This new version of PAM has supervised principal components for predicting survival (or a quantitative outcome) from gene expression data.
PAM uses R functions written by Trevor Hastie and Rob Tibshirani.
For classification problems, PAM implements the 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 problems, PAM implements the supervised principal components method:
Semi-supervised methods for predicting patient survival from gene expression papers
(Bair, Tibshirani) PLOS
Prediction by supervised principal components (Bair, Hastie, Paul, Tibshirani) Stanford tech report
A number of bugs were fixed, and compatibility with R1.9.1 was achieved. The corresponding R package pamr 1.21 has also been updated, and can be used as a standalone package in R.
The installation procedure for PAM for Excel (version 1.20) has been greatly simplified. The software can be now installed and uninstalled like a standard windows application. In addition, a new and improved imputation engine has been implemented, extensions to Survival Analysis problems have been introduced and some bugs fixed.
PAM for Excel requires the R package and Microsoft Excel. The installation process checks for R. If you don't have R, you first need to install a recent version of the R statistical package. This is free, and can be found at : The R website. Follow the instructions. Click on CRAN under Download. Click on Windows (95 or later), click on base/ and download the latest version of the R executable (currently rw1091.exe). Installation takes 5-10 minutes.
If you installing PAM for Excel for the first time, the procedure is very simple. Just download the executable and double click on it to set it up. PAM will be installed and the PAM buttons will automatically appear the next time you fire up Excel.
If you are upgrading from a previous version of PAM, then the procedure is, unfortunately, a bit more involved. Of course, such an involved process need be followed only once as subsequent upgrades will be simple.
You need administrative privileges.
PAM. Exit Excel.
remove.packages("pamr"). Exit R.
Now double click on the executable you downloaded to install PAM for Excel like a standard windows application.
You're now ready to use PAM.
C:\Program Files\PAMVB\Examples) and double click on
khan.xls. This is an example with 4 cancer classes and 2308 genes.
Notice that the data setup for PAM is much like that for SAM. There is one line per gene. There are additional rows at the top of the spreadsheet for class labels and (optional) sample and batch labels.
Class Labels in selection row: 2
Sample Labels in selection row: 1
Batch Labels in selection row: leave blank
Expression Data starts in selection row: 3
Click on OK to start the PAM computation.
Click on Train and other active buttons to get results. Of course, you must choose the threshold for shrinkage to do the more interesting things.