Analyses were corrected for false discovery rate.
The Benjamini-Hochberg method was used to determine differentially expressed transcripts and control for false discovery rate.
After a correction for false discovery rate at 10%, 72 genes were found to be significantly up-regulated (40) or down-regulated (32).
Adjusting for false discovery, 14 analytes including MCP1c, IP-10, Leptin, interleukin (IL -5, EotaxIL -5L-6, G-CSF, CXCL5 changEotaxinificantly post TPE induction.
Given the reduction in the number of variables, the variable cluster analysis approach proved to be more powerful when correcting for false discovery rate.
Data were analyzed using the exact (permutation) version of the Chi-square test in R [40], and checked for false discovery rate according to Benjamini et al. [41].
All p-values shown are Benjamini-Hochberg corrected for false discovery, unless otherwise noted.
A hypergeometric test with subsequent correction for false discovery rate (FDR) when using multiple testing was applied.
It was then searched again using Mascot to generate the protein list for false discovery rates (FDR) evaluation.
P-values were adjusted for false discovery using the Benjamini-Hochberg procedure [35] or q-value method [36] where applicable.
After controlling for false discovery rate [21], neither significant linkage disequilibrium nor departures from HW were detected with the allozyme data.
