Hi all, I am trying to use the SNM package to normalize a dataset. The whole dataset is a combination of Illumina arrays and Affymetrix arrays. I use the batch effect to model the platform difference. Following is my code for constructing the bio.var and adj.var matrices. The SNM function gives an error of invalid model matrix. I noticed that on the SNM help page, it highlights the necessity of making the model matrix valid. However, I am not sure how to determine if the model matrices are valid. Could anybody please help? Thank you very much. Wendy > Bio_var<-model.matrix(~groups,BioNames) #65x12 matrix > Adj_var<-model.matrix(~batches,Batch) # 65x1 matrix > snm.obj1<-snm(raw.data,bio.var=Bio_var,adj.var=Adj_var,int.var=NULL) Error in make.snm.obj(Y = raw.dat, bio.var, adj.var, int.var, spline.dim, : cbind(bio.var,adj.var) is not a valid model matrix. Enter '?model.matrix' for more information on building a model matrix. [[alternative HTML version deleted]]