Fig. 9

Fertility in an independent cohort can be predicted from DMCs. A a Random Forest model was built on the main cohort using the DNA methylation values of 107 fertility-related DMCs without missing values. The model was then run on the independent cohort, and the predicted fertility status was compared with the actual fertility status. ROC curves, accuracy, AUC, sensitivity and specificity were comparable to the performance indicators obtained when the testing set was included in the main cohort. B principal component analysis was run on the 107 fertility-related DMCs, and an individual factor map is displayed, with bulls colored according to the consistency between their predicted and actual fertility status (red: fertile predicted as fertile, FF; blue: subfertile predicted as subfertile, SS; orange: fertile predicted as subfertile, FS). Confidence ellipses are indicated. Misclassified bulls were found in the fertile class only and displayed a DNA methylation pattern similar to that of the subfertile class