[BioC] Question about topTable

Mark Cowley m.cowley at garvan.org.au
Mon Sep 27 01:34:44 CEST 2010


Hi Casper,
logFC is the difference between two groups, so it is not defined when you do an F-test which in essence compares any number of groups. If you do have some specific comparisons of interest, then specify the coef argument when you run toptable
eg toptable(my.fit, coef=1, ...)

When you choose method="BH", toptable just calculates the FDR, using the BH method, and puts the FDR values in the adj.p.val column. toptable itself does no thresholding on those values. You choose the threshold yourself. So if you choose only those genes where adj.p.val < 0.05, then you have controlled the FDR to a level of 5%

hope that helps

Mark

On 25/09/2010, at 6:54 AM, Casper Shyr wrote:

> 
> Dear BioC,
> 
>  According to R documentation for toptable function, the parameter lfc doesn't seem be available for topTableF. Does this mean if I am to use toptableF to view DE genes for a set of contrasts, I am not able to filter genes based on their log-fold change?
> 
> 
> 
>  Also, on a related note, under p.adjust argument, in the description for "BH" method, it said this method controls FDR to be below a specified value. How do I find out what value this is taking by default, and also, how would I go about changing it?
> 
> 
> 
> Thank you!
> 
> Sincerely,
> 
> Casper
> 		 	   		  
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor



More information about the Bioconductor mailing list