Hi Everyone,
I have been using EGSEA recently on my RNA seq datasets with no issues. Recently, I suggested a student try it with their Affymetrix data and hit a snag. We are pre-processing with RMA through minfi, filtering and using that as the expression matrix. We then generate a probe.annotation data frame including the probe IDs and entrez IDs from the array annotation file. Design and contrast matrices and the group vector are specified as suggested by the authors. When running the egsea.ma function we get the following error...
Error in dimnames(x[[a]]) <- value :
length of 'dimnames' [1] not equal to array extent
I am failrly certain the problem is with the expression matrix or probe.annotation frame since the error comes up with this stripped down version...
egsea.ma(expr=RMA, group=t, probe.annot = probe.annot)
I can easily run this function with either of the examples the authors provide and can't for the life of me figure out how those objects differ from the ones I am trying to use. If anyone knows where in the source code this error is likely occuring or knows what the check is for, please let me know. I realize I haven't provided any reproducible code to work off of, so let me know what I can post to be more helplful.
Many thanks!
UPDATE
I figured out the issue.... I had too many NA and duplicated Entrez IDs in my probe.annotation object. Once I removed these and trimmed my expression matrix accordingly, everything ran fine. I should have thought of that in the first place!
