[BioC] Fisher-test

Naomi Altman naomi at stat.psu.edu
Sat Aug 7 05:41:26 CEST 2010

You are not using the Fisher test correctly.  The entries in the 
table should be counts.


At 05:00 PM 8/3/2010, Noe Andor wrote:
>I have a question regarding fiher-test & differential gene 
>expression. I usually
>use ttest to see which genes are differential expressed between two
>sample-types. I thought of using fisher test for the first time, to see if
>results are comparable, yet I'm not shure if I applyed it correctly. I made a
>2x2 contingency table for each gene, like:
>    Notch1-Normal    Notch1-Tumor
>up            x                        x2
>down        y                        y2
>x = Sum of all foldchanges from normal samples (min foldchange = 1.5)
>y = Sum of all foldchanges from normal samples (max foldchange = 0.5)
>...analogue for tumor samples.
>Then I applyed Fisher test:
>tab<-c(x, x2, y, y2)
>fisher.test(tab, alternative="two.sided")
>Yet I got no significant results (as compared to the t-test results, which
>harbor some good hits) - so I assume I did'n built the table the right way.
>I would be greatfull for an example, especially regarding that contingency
>best regards,
>         [[alternative HTML version deleted]]
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>Search the archives: 

Naomi S. Altman                                814-865-3791 (voice)
Associate Professor
Dept. of Statistics                              814-863-7114 (fax)
Penn State University                         814-865-1348 (Statistics)
University Park, PA 16802-2111

More information about the Bioconductor mailing list