A largely used concept of error control in such multiple testing is the expected proportion of falsely rejected hypotheses, or False Discovery Rate (FDR).
Univariate analysis of metabolomics data usually includes hundreds of tests (one for each metabolite), therefore the control of false discovery rate (FDR) in multiple testing is very important.
Furthermore, if a Bonferroni correction for multiple testing is applied, none of the SNPs would be individually significantly associated with MM.
Additionally, the authors showed that adjustment for multiple testing is mandatory to derive potential biomarkers that can be verified in an independent test set.
The central concern for multiple testing is addressed by three features of the study.
The only other significant pathway after adjusting for multiple testing is MAPK signaling, which is downregulated in the 'D' group.
This error rate can be much higher than the significance level if no correction for multiple testing is applied.
Since multiple lineages were evaluated this way, multiple testing is a potential problem for inferring the significance of results.
Additionally, the high number of tests to be examined reduces statistical power dramatically when correction for multiple testing is taken into account.
Finally, PGMDR results – and in fact also GMDR results – are too liberal as no correction for multiple testing is carried out.
Finally, a correction for multiple testing is applied to the p-values, and proteins with significant changes in abundance are determined.
