b4303a4

committed
# Commits

# Comments (0)

# Files changed (1)

# tools/rgenetics/rgedgeRFactorial.xml

help="May be a good or a bad idea depending on the biology and the question. This was the old default. Quantile based is available as an alternative"/>

<param name="priorn" type="integer" value="4" size="3" label="prior.df for tagwise dispersion - lower value = more emphasis on each tag's variance - note this used to be prior.n"

help="Zero = auto-estimate. 1 to force high variance tags out. Use a small value to 'smooth' small samples. See edgeR docs and note below"/>

- <param name="fdrthresh" type="float" value="0.05" size="5" label="P value threshold for FDR filtering for amily wise error rate control"

+ <param name="fdrthresh" type="float" value="0.05" size="5" label="P value threshold for FDR filtering for family wise error rate control"

edgeIt = function (Count_Matrix,group,outtab1,outtab2,outtab3,fdrtype='fdr',priorn=5,fdrthresh=0.05,outputdir='.',

print.noquote("Number of conditions identified in experiment does not equal 4 - full 2x2 factorial not possible")

-Performs digital gene expression factorial analysis between two factors - typically a treatment applied to 2 different cell types, or two treatments applied in a factorial design.

-Data with the four groups is supplied as a count matrix and the two primary comparisons are defined by selecting samples to represent

-control and treatment (whatever that means) for each comparison. The interaction is then defined as the difference between those two

-comparisons. All comparisons are reported as separate tabular spreadsheets ordered by p value and a comprehensive summary is

+Like the pairwise version, it takes a count matrix with columns containing reads per contig (NOT transformed!) for

+multiple replicates in comparison groups. The counts can be any kind of counts. It was designed for

+Factorial designs are old but good if you want to get at the individual effects and interaction between two factors.

+EG: a treatment applied to 2 different cell types, or two treatments applied in all 4 possible combinations (-a-b,-a,+b,+a-b,+a+b).

+The interaction is defined as the difference between those two comparisons and is reported as a topTable as are the

+All comparisons are reported as separate tabular spreadsheets ordered by p value and a comprehensive summary is

+This code essentially embelishes the code described by Gordon Smythe in the limma documentation for a factorial

+Tabular files which contain the statistical results and the raw and transformed counts and some colourful

to do because it wasted hours of my time to track down and will similarly cost other edgeR users dearly

so you could consider using a smaller prior.n, although I would hesitate to use a prior.n less than 5.

+ .. _edgeR_Manual: http://www.bioconductor.org/packages/release/bioc/vignettes/edgeR/inst/doc/edgeRUsersGuide.pdf