Figure 2

Prediction accuracy using sequencing data. Prediction results for the simulated phenotype and real phenotype using sequencing data. We ran 10-fold cross-validation 20 times to generate these boxplots. The x-axis is the index of the regularization parameter $\lambda$. The corresponding number of SNP markers used in the prediction model is given in brackets. Notice that these numbers are obtained using all samples. Index 0 corresponds to prediction only using pedigree information, thus the corresponding number of SNP marker is zero. The y-axis is the R² between the true value and the prediction value.