metabolomics-statistics

To identify differentially abundant metabolites, apply univariate statistical tests like Welch's t-test or ANOVA to your metabolomics data, then use FDR correction to control for false discoveries.

Available metabolomics statistical tests include Welch's t-test, Wilcoxon rank-sum, ANOVA, and Kruskal-Wallis, enabling robust differential analysis between biological groups.

You can run univariate tests on a normalized metabolites CSV file by using pandas for data manipulation and scipy for statistical computations, saving the biomarker discovery results to an output directory.

FDR correction, specifically Benjamini-Hochberg, is needed for high-dimensional metabolomics data to control false discoveries when performing multiple univariate tests across many metabolites.

Yes, you need normalized metabolomics data before applying statistical tests, as the quick start process explicitly requires running Welch's t-test on a normalized metabolites data file.

Yes, you can use pandas and scipy for metabolomics data manipulation and statistical computations, as they are the required dependencies for performing univariate tests and FDR correction.