hi, i am using limma to analyze agilent chip data with four groups of cell lines: wt, mut, wt+S(stimulated), mut+S. goals: 1)find genes difference between wt and mut; 2)which genes response to S in wt and mut repectively; 3) which genes respond differently in mut and wt. can anyone tell me how to design a matrix model for the comparison ? thanks. yong

On 1/10/06 2:19 AM, "Yong Jiang" <yjiang at="" ucsd.edu=""> wrote: > hi, > i am using limma to analyze agilent chip data with four groups of cell > lines: wt, mut, wt+S(stimulated), mut+S. > goals: 1)find genes difference between wt and mut; 2)which genes > response to S in wt and mut repectively; 3) which genes respond > differently in mut and wt. > > can anyone tell me how to design a matrix model for the comparison ? > thanks. This is a 2x2 factorial design. There are good examples in the limma user guide. The weaver mutant data are from a two-color 2x2 factorial design. That section should be particularly helpful. Sean