Fig. 1
Optimal data dichotomization for antibody selection. The different steps of the analysis are displayed on the workflow using distinct colored shapes. Blue color identifies the beginning of the pipeline where the antibody values are sorted. Light orange identifies the loop for obtaining the χ2 test p-values for each potential cut−off. Green indicates the selection of the most significant cut-off. Dark orange refers to the assessment of the statistical significance of the most significant cut-off after controlling for the False Discovery Rate (FDR) with the Benjamini−Yekutieli procedure. Red refers to the implementation of the Super Learner and the computation of the classification probability. Additional information is provided by the faded light orange and red colored shapes