EdgeR multi-factâ or testing questions
guest at bioconductor.org
Wed Jan 8 14:39:04 CET 2014
Thank you very much for your comments, I really appreciate it.
Definitely we need to consider the principal of marginality. Previously I thought that we can fit one model including everything, i.e., main effects, all two-interaction terms and the three-way interaction term, then use glmLRT to test all these terms using the crresponding coefficients in the model, and the pvalue results of these glmLRT tests will be the same as anova() function where the pvalue is reported for each term.
I will fit the different models to test the main effects and the two-way interaction terms based on the significance results of high-level interaction term.
-- output of sessionInfo():
R version 3.0.1 (2013-05-16)
Platform: x86_64-w64-mingw32/x64 (64-bit)
 LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252 LC_MONETARY=English_United States.1252
 LC_NUMERIC=C LC_TIME=English_United States.1252
attached base packages:
 stats graphics grDevices utils datasets methods base
other attached packages:
 edgeR_3.2.4 limma_3.16.8
loaded via a namespace (and not attached):
Sent via the guest posting facility at bioconductor.org.
More information about the Bioconductor