While doing exactTest in EdgeR,
I can choose an option, dispersion= "common", "tagwise", or "auto".
My question is what is the difference between these two options. Which option is appropriate?
In my data analysis, common.dispersion 0.3064238.
If I did exactTest with dispersion="common", I've got p-values below.
However, if I did exactTest with dispersion="tagwise", I've got p-values below.
These two results look quite different. Since I wanna use B&H (1995) method for p-value adjustment, if I choose common dispersion, I would gain a list of differentially expressed genes. However, if I choose tagwise dispersion, I would gain no gene which is differentially expressed.
Thank you in advance.
I can choose an option, dispersion= "common", "tagwise", or "auto".
My question is what is the difference between these two options. Which option is appropriate?
In my data analysis, common.dispersion 0.3064238.
If I did exactTest with dispersion="common", I've got p-values below.
PValue (sorted from smallest to largest)
8.34E-08
1.23E-07
3.30E-07
3.81E-07
3.94E-07
4.72E-07
5.22E-07
5.27E-07
5.44E-07
8.34E-08
1.23E-07
3.30E-07
3.81E-07
3.94E-07
4.72E-07
5.22E-07
5.27E-07
5.44E-07
PValue (sorted from smallest to largest)
0.00393925
0.003999847
0.004926992
0.005319714
0.005459274
0.005504967
0.005744749
0.005883501
0.005897807
0.00393925
0.003999847
0.004926992
0.005319714
0.005459274
0.005504967
0.005744749
0.005883501
0.005897807
Thank you in advance.