Hi I did an analysis for some scientist in April using limma and knn.impute. I did the same analysis on the same data set and the same commands yesterday but I got a different result. More genes were classified as differential expressed and some of the genes that were differential expressed earlier are not any more. I think that the reason for this might be that R and the Bioconductor packages has been updated since the last time I run these commands on this data set. If this is the reason for the different results are the most trustworthy, and how do I explain this to the scientist without them loosing their faith in R/Bioconductor? If the changes in R and Bioconductor can not cause these changes in the results, what might be the reason? Regards, Ingrid -- Ingrid H.G. ?stensen Bioinformatics user support (NMC) Department of Tumour Biology The Norwegian Radiumhospital 0310 Oslo E -mail: ingrid at microarray.no Phone: +47 22 93 54 11 Fax: +47 22 52 24 21

Hi Ingrid, "Ingrid H.G. ?stensen" <ingrid at="" microarray.no=""> writes: > I did an analysis for some scientist in April using limma and knn.impute. > I did the same analysis on the same data set and the same commands > yesterday but I got a different result. More genes were classified as > differential expressed and some of the genes that were differential > expressed earlier are not any more. Without more details it is difficult to provide a meaningful answer. Some details that would be useful include R and BioC package versions before and after as well as further details of what you did that is giving different results. It could be a bug. It could be a bug that was fixed. It could be a change in default behavior. It could be related to changes in annotation data. + seth