[<i>2-column fit table]</i> r² and m_rmse values with standard deviation of 100 bootstrap runs for all models.

For Random Forest and Elastic Net ‘All Genes’ summary, the ‘Full Models’ use all genes from data (32303) whereas the ‘Reduced Models’ use only the top 15 genes from the full model. The Random Forest Literature full model was done with 85 genes referenced in S7 Table in S4 File, and the reduced model was run with only the top 15 genes from this subset. The models run for qPCR used only the 15 genes selected for qPCR, therefore a reduced model was not applicable. Performance metrics are reported with ± standard deviation.