Multiple tests, one per row

In case of qPCR, or small scale transcriptomics, proteomics or metabolomics, you sometimes need to do one test per row (rows being genes, proteins or metabolites). The test identifies transcripts, proteins or metabolites whose abundance changes in a group of treated samples compared to a group of control samples. It is crucial to perform a multiple testing correction to avoid generating a massive amount of false positives.

Various tests, one per row

Overview of various tests comparing 2 or more than 2 groups for each row (gene, protein, metabolite) in the data set. Note that the data should be a data frame in long format with a column that contains the grouping info.

• R script for Multiple comparisons, one per row
• data set for Multiple comparisons, one per row

Note that for large scale omics data, specialized packages (like DESeq2, MSQRob,… ) exist that do a better job analyzing the data than the approach described above because they offer a better handling of the peculiarities of the data.