Fig. 1

Flowchart describing the pipeline applied for machine learning-based predictions for residual feed intake (RFI) and the feed conversion ratio (FCR) using the mean methylation level within differentially methylated regions (DMRs) and genetic variants (single-nucleotide variants and insertions and deletions). The models in the hyperparameter tuning step were selected on the basis of the Spearman correlation (rho). After this step, the best overall model was selected on the basis of the ratio between the squared Spearman correlation (rho²) and the root mean squared error (RMSE)