A Deep Learning Approach for Predicting Antigenic Variation of Influenza A H3N2

Table 2

The prediction results obtained from different deep learning models with the position feature alone and in combination with the other three features.

Model	Agreement (%)	Sensitivity (%)	Specificity (%)	MCC
Position	95.73	95.18	96.12	0.914
Position-epitope	97.16	96.85	97.34	0.939
Position-Gly	95.02	93.84	95.75	0.895
Position-RBD	94.74	92.42	96.44	0.892