Hi! All, I am trying to analyze very large microarray data sets with 550 arrays. But there are lots of missing values (NA) in the data set which is making my life difficult. So I need to get rid of those NAs in the data. I do not want to impute. Also, I can not remove those rows containing NAs as almost each row have it. Is it a good idea to replace NAs with zero in the data set. If yes, why? and if not, why not? Your suggestion will be invaluable. cheers! Amit