Figure EV4. Batch effects in the Bariatric surgery study.

(A) Principal Variance Component Analysis shows a clear effect of the liquid handling robot’s channels (robot_batch (plate rows A and B vs. C–F)). (B) This technical factor is efficiently removed after ComBat batch correction. (C) Distribution of correlations between and within batches for different time points from the same animal and different animals prior to and (D) post‐batch correction. Peptide intensities in C and D are normalized to pre‐surgery intensity values. The procedure improves the correlation of samples coming from the same animal at different time points which can be used as pseudo‐biological replicates in this study.