I have been using the affy package to summarize and filter .cel files into expression sets, then exporting these sets into SAM (Significance Analysis of Microarrays) to select differentially expressed genes. This is getting cumbersome; it would be more straightforward to do everything in R. Does anyone have functions for SAM in R, or can I use any of the existing bioconductor packages to select differentially expressed genes in a way similar to SAM? Thanks Thomas Cappola, M.D. Assistant Professor of Medicine Heart Failure and Transplantation University of Pennsylvania School of Medicine 6 Penn Tower 3400 Spruce Street Philadelphia, PA 19104 Tel 215-615-0805 Fax 215-615-0828 thomas.cappola@uphs.upenn.edu [[alternate HTML version deleted]]

> I have been using the affy package to summarize and filter .cel files into > expression sets, then exporting these sets into SAM (Significance Analysis > of Microarrays) to select differentially expressed genes. > This is getting cumbersome; it would be more straightforward to do > everything in R. > Does anyone have functions for SAM in R, or can I use any of the existing > bioconductor packages to select differentially expressed genes in a way > similar to SAM? hi Thomas, and others, See the announcement below by Holger Schwender -- the R code works fine. Stephen ---------------------------------------------------------------------- because of the large number of requests I uploaded my functions that can perform the SAM analysis and the Empirical Bayes Analysis described in the JASA-Paper of Efron to the internet. You can download the S-Plus functions from http://de.geocities.com/holgerschw/GENE-SIG.ZIP A short description you will find under http://de.geocities.com/holgerschw/gene-sig.pdf An older version of the R functions is available at http://de.geocities.com/holgerschw/r.zip These R functions cannot handle a very large number of permutations (let's say B larger than 500 or 1000). There is a newer version of these functions that can handle this number of permutations. I will post these functions as soon as I am back home (Sunday or Monday).

You will not be able to download the scripts from that page any longer. Holger took them down after being counseled that he may be in legal trouble given that there are patents pending on the SAM algorithms. My understanding is that he contacted Rob Tibshirani for permission to post the scripts but has not received an answer. Jim James W. MacDonald UMCCC Microarray Core Facility 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623 >>> Stephen Eglen <stephen@inf.ed.ac.uk> 06/04/03 05:41AM >>> > I have been using the affy package to summarize and filter .cel files into > expression sets, then exporting these sets into SAM (Significance Analysis > of Microarrays) to select differentially expressed genes. > This is getting cumbersome; it would be more straightforward to do > everything in R. > Does anyone have functions for SAM in R, or can I use any of the existing > bioconductor packages to select differentially expressed genes in a way > similar to SAM? hi Thomas, and others, See the announcement below by Holger Schwender -- the R code works fine. Stephen ---------------------------------------------------------------------- because of the large number of requests I uploaded my functions that can perform the SAM analysis and the Empirical Bayes Analysis described in the JASA-Paper of Efron to the internet. You can download the S-Plus functions from http://de.geocities.com/holgerschw/GENE-SIG.ZIP A short description you will find under http://de.geocities.com/holgerschw/gene-sig.pdf An older version of the R functions is available at http://de.geocities.com/holgerschw/r.zip These R functions cannot handle a very large number of permutations (let's say B larger than 500 or 1000). There is a newer version of these functions that can handle this number of permutations. I will post these functions as soon as I am back home (Sunday or Monday). _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor

Thanks for your replies. If the code is reposted, please let us know! Tom Thomas Cappola, M.D. Assistant Professor of Medicine Heart Failure and Transplantation University of Pennsylvania School of Medicine 6 Penn Tower 3400 Spruce Street Philadelphia, PA 19104 Tel 215-615-0805 Fax 215-615-0828 thomas.cappola@uphs.upenn.edu -----Original Message----- From: James MacDonald [mailto:jmacdon@med.umich.edu] Sent: Wednesday, June 04, 2003 7:23 AM To: stephen@inf.ed.ac.uk; thomas.cappola@uphs.upenn.edu Cc: bioconductor@stat.math.ethz.ch Subject: Re: [BioC] SAM for R You will not be able to download the scripts from that page any longer. Holger took them down after being counseled that he may be in legal trouble given that there are patents pending on the SAM algorithms. My understanding is that he contacted Rob Tibshirani for permission to post the scripts but has not received an answer. Jim James W. MacDonald UMCCC Microarray Core Facility 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623 >>> Stephen Eglen <stephen@inf.ed.ac.uk> 06/04/03 05:41AM >>> > I have been using the affy package to summarize and filter .cel files into > expression sets, then exporting these sets into SAM (Significance Analysis > of Microarrays) to select differentially expressed genes. > This is getting cumbersome; it would be more straightforward to do > everything in R. > Does anyone have functions for SAM in R, or can I use any of the existing > bioconductor packages to select differentially expressed genes in a way > similar to SAM? hi Thomas, and others, See the announcement below by Holger Schwender -- the R code works fine. Stephen ---------------------------------------------------------------------- because of the large number of requests I uploaded my functions that can perform the SAM analysis and the Empirical Bayes Analysis described in the JASA-Paper of Efron to the internet. You can download the S-Plus functions from http://de.geocities.com/holgerschw/GENE-SIG.ZIP A short description you will find under http://de.geocities.com/holgerschw/gene-sig.pdf An older version of the R functions is available at http://de.geocities.com/holgerschw/r.zip These R functions cannot handle a very large number of permutations (let's say B larger than 500 or 1000). There is a newer version of these functions that can handle this number of permutations. I will post these functions as soon as I am back home (Sunday or Monday). _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor [[alternate HTML version deleted]]