Performance of integrated training sets.

A) Comparison of the performance of non-integrated training sets with integrated training sets in ECO. The boxplot with X axle 1 indicates the AUC score distribution with non-integrated training sets. The boxplots with X axles 2, 3, and 4 indicate the AUC score distributions with the integrated training sets, where 2, 3, and 4 represent integration of 2, 3, and 4 species, respectively, out of the rest of the 19 species. B) Comparison of the performance of non-integrated training sets with integrated training sets in SCE. C) Comparison of the difference in diverse groups of integrated training sets. The blue histogram refers to low AUC group, the green histogram refers to mixed AUC group, and the red histogram refers to high AUC group. The blue dashed line indicates the upper bound of LASTS, and all training sets in LASTS have lower AUC score than this value. The red dashed line indicates the lower bound of HASTS, and all training sets in HASTS have higher AUC score than this value.