The FDR is an adjustment of p values where the adusted p values are larger than the (raw) p values taking into account multiple testing. The classical FDR was introduced by

Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B 57, 289–300.

In recent years many alternative approaches to control FDR have been proposed; see for instance http://www.amazon.com/Multiple-Procedures-Applications-Genomics-Statistics/dp/1441923799

Regarding Matlab see http://de.mathworks.com/help/bioinfo/ref/mafdr.html

Best

Matthias

See the following resources as well.  There are two PDF's on how to do it and one excel file that will do it for you if you enter your output. 

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CCoQFjAAahUKEwih-LyP9orGAhXMFpIKHX_bBXA&url=http%3A%2F%2Fwww.air.org%2Fsites%2Fdefault%2Ffiles%2Fdownloads%2Freport%2FWWC_Phase_I_Computation_Tools-4-15-10_0.xls&ei=Yzd7VeG6M8ytyAT_tpeABw&usg=AFQjCNHQbtxbx3qqT-uyPla-GjwNLwJNFQ&sig2=h9MO-a4n1ge_8Bjuao3CPg&bvm=bv.95515949,d.aWw

Regarding the FDR-adjusted p-value here's a formula:

i=Ranked p value (eg. 1-100)

m=Number of tests e.g. (1-1000)

Q= The false discovery rate (5%-25%)

Formula: (i/m)Q=FDR-adjusted p-value (can also be called q value).

Here’s two websites that explain it well:

http://www.nonlinear.com/support/progenesis/comet/faq/v2.0/pq-values.aspx

https://brainder.org/2011/09/05/fdr-corrected-fdr-adjusted-p-values/

These are covered in more detail in Bejamini-Hochberg's 1995 and 1998 papers.

I used the Benjamin-Hochberg procedure in a Pearson correlation analysis stating the FDR for all correlations in a manuscript I recently submitted. I got the manuscript back with a reviewer comment saying "I recommend stating the P values of correlations shown in fig 2 and 3 which the authors mention to be highly significant". I am at odds how to write a rebuttal. Any tips?