l*****k
发帖数: 587 | 1 again, here is some misunderstanding about fold change and P value.
fold change between two chips means nothing, since it could be random or due
to bad probeset or what ever error message.
P value means more, since it is calculated from statistics principle (in most
cases, they, wrongly, assume normal distribution of the gene expression data.
that is why you have to do log transformation, which makes things more like
normal distributed)
a very consistent change, even with small fold change value |
|
