Dear Matthew, I guess the main publication is my own - at present only an unrefereed conference proceeding. http://www.stat.psu.edu/%7Ewzhao/bcc/respapers/AltmanInterface04.doc I think that the Kerr and Churchill papers do it this way, too, but they do not explain why. Basically, using differences has 2 disadvantages - it has higher variance than single channel analysis, and if a spot is missing on an array, many of the effects for that gene cannot be estimated. At 04:07 PM 1/3/2006, you wrote: >Dear Naomi, > >I recently read a post of yours on the bioconductor listserv that >said you were a "big believer in single channel analysis of loop >designs." Can you point me towards any publications that sway your >opinion that way? I've always done dual channel analysis of loop >designs and never really thought much about the distinction. > >Thanks in advance for your help. > >Matt > >-- >Matt Scholz >Senior Research Specialist >Department of Plant Sciences >University of Arizona >(520) 621-1695 Naomi S. Altman 814-865-3791 (voice) Associate Professor Dept. of Statistics 814-863-7114 (fax) Penn State University 814-865-1348 (Statistics) University Park, PA 16802-2111

Dear Matthew and Naomi: You may be interested also on taking a look at: G. J. M. Rosa, J. P. Steibel, R. J. Tempelman (2005) Reassessing design and analysis of two-colour microarray experiments using mixed effects models. Comparative and Functional Genomics 6(3): 123-131. http://www3.interscience.wiley.com/cgi-bin/jissue/110473749 for some discussion on different levels of replication (technical and biological), and appropriate modeling of such data structures. All the best, Guilherme. > Dear Matthew, > I guess the main publication is my own - at present only an > unrefereed conference proceeding. > http://www.stat.psu.edu/%7Ewzhao/bcc/respapers/AltmanInterface04.doc > I think that the Kerr and Churchill papers do it this way, too, but > they do not explain why. > > Basically, using differences has 2 disadvantages - it has higher > variance than single channel analysis, and if a spot is missing on an > array, many of the effects for that gene cannot be estimated. -- Guilherme J. M. Rosa Assistant Professor Department of Animal Science Department of Fisheries and Wildlife Michigan State University 1205-I Anthony Hall East Lansing, MI 48824-1225 USA Phone: + 1 517 353-5102 Fax: + 1 517 353-1699 E-mail: rosag at msu.edu http://www.msu.edu/~rosag/