Multiple-testing correction in metabolome-wide association studies

Abstract Background The search for statistically significant relationships between molecular markers and outcomes is challenging when dealing with high-dimensional, noisy collinear multivariate omics data, such as metabolomic profiles. Permutation procedures allow the estimation of adjusted significance levels without assuming independence among variables. Nevertheless, complex non-normal structure metabolic profiles may bias permutation results leading to overly conservative threshold estimates i.e. lower than those from a Bonferroni or Sidak correction. Methods Within univariate procedure we employ parametric simulation methods based on (log-)Normal distribution obtain which are consistent across different while effectively controlling type I error rate. Next, derive an alternative closed-form expression number non-redundant variates spectral decomposition their correlation matrix. performance method tested model parametrizations wide range using synthetic real data sets. Results Both permutation-based formulation more practical closed form found give effective indication independent effects exhibited by system, guaranteeing that derived stable outcome measures diverse properties.