[BioC] how to design a model matrix and test by edgeR

[wang peter]

dear Dr Gordon:
          thank you very much for your reply.
for the first question, it should be cof=14.
i made a mistake.

and for the dispersion=dge$common.dispersion
i make a mistake either.

but thank you for youre corrrection.

another question, if you donot mind.
do i need estimate  the common dispersion
before i estimate the tagwised dispersion

thank u
shan