Table 2.

AUC performance of SVM classifier on embedding features with different biological word lengths

Feature	AUC
	5-fold cross-validation	Independent
1-g	0.856 ± 0.485	0.848 ± 0.525
2-g	0.901 ± 0.416	0.883 ± 0.599
3-g	0.934 ± 0.423	1 ± 0
4-g	0.617 ± 0.63	0.563 ± 0.751
5-g	0.543 ± 0.539	0.574 ± 0.686
1-g and 2-g combined	0.952 ± 0.416	0.934 ± 0.48
1-g and 3-g combined	0.96 ± 0.42	0.921 ± 0.497
2-g and 3-g combined	0.984 ± 0.298	0.998 ± 0.277

(Each result is reported in format: m ± d, where m is the mean and d is the standard deviation across the ten runs)