5.3 years ago by
Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia
The Shapiro test is only applicable to iid samples, so it is difficult
see how it could be used to test normality of expression values in a
linear modelling context. If you have applied the test to the
expression values for each gene, then I suspect that the test is
picking up differential expression rather than non-normality.
The limma code is very robust against non-normality. All the usual
microarray platforms and standard preprocessing procedures produce
that is normally distributed to a good enough approximation. Much
has been devoted to developing good preprocessing and normalization
The concept of "robustness" in statistical analysis goes back a 1953
by George Box in Biometrika. In that paper, Box wrote of the
property of robustness to non-normality which [tests for comparing
possess". The tests done by limma inherit the robustness property
Box was referring to. Box made the point that the robustness of the
sample t-test was not improved by checking first for equal variances.
"To make the preliminary test on variances is rather like putting to
in a rowing boat to find out whether conditions are sufficiently calm
an ocean liner to leave port!"
I rather think that, if Box was still alive today, he might say
similar about a preliminary Shapiro test!
> Date: Thu, 21 Nov 2013 17:42:21 -0500
> From: Jerome Lane <jerome.lane at="" criucpq.ulaval.ca="">
> To: "bioconductor at stat.math.ethz.ch" <bioconductor at="" stat.math.ethz.ch="">
> Subject: [BioC] [limma] [Rfit] [samr] Gene expression distribution
> using lmFit and eBayes
> The 3/4 of my microarray gene expressions have non normal
> most of p-values after Shapiro test under 10x-5.
> I tried linear ranked regression from rfit (no normality
> residues) from Rfit package for adjustment of covariables + SAM
> parametric) from samr package but results where not as
> as lmFit + eBayes could provide.
> I know that lmFit function can analyses gene expression not
> but what is the limit ?
> Is it statistically relevant to use lmFit + eBayes according to my
> Best regards,
> Jerome Lane
The information in this email is confidential and