Identification of plant vacuole proteins by using graph neural network and contact maps

BMC Bioinformatics

Table 5 The performance of models on the independent test set

Model	Acc (%)	F1-score	Sp (%)	Sn (%)	MCC	ROC-AUC	PR-AUC
GaussianNB	63.51	0.6667	54.05	72.97	0.2752	0.6963	0.6971
LR	72.97	0.7500	64.86	81.08	0.4656	0.7509	0.7111
RF	71.62	0.7308	66.22	77.03	0.4350	0.7669	0.7560
SVM	69.59	0.6939	70.27	68.92	0.3919	0.7763	0.6976
LightGBM	68.24	0.6846	67.57	68.92	0.3649	0.7431	0.7100
GBDT	64.86	0.6667	59.46	70.27	0.2990	0.7162	0.7128
MLP	64.87	0.6338	68.92	60.81	0.2983	0.7476	0.7339
KNN	60.14	0.6335	51.35	68.92	0.2059	0.6443	0.6186
XGBoost	65.54	0.5641	86.49	44.59	0.3423	0.6883	0.6824
GraphIdn	88.51	0.8917	82.43	94.59	0.7760	0.9326	0.9140

ISSN: 1471-2105