Jenner-predict server: prediction of protein vaccine candidates (PVCs) in bacteria based on host-pathogen interactions

Table 1 Performance evaluation of Jenner-Predict server against existing software, NERVE, and web servers, Vaxign and VaxiJen ^*

S. No.	Different software and web servers	Prediction in	Prediction in	^$Protegen (bacterial)		^%Data used for VaxiJen server development
		S. pneumoniae ^#	E. coli ^#	No of predictions	Sensitivity	Positive data	Negative data	Sensitivity (Specificity)
1.	NERVE	7 (18) [58]	21 (28) [527]	121 (177)	0.684	53 (83)	8 (33)	0.639 (0.758)
2.	Vaxign	6 (18) [61]	18 (28) [286]	89 (177)	0.502	41 (83)	5 (33)	0.494 (0.848)
3.	VaxiJen	3 (18) [313]	21 (28) [955]	97 (177)	0.548	46 (83)	3 (33)	0.554 (0.909)
4.	Jenner-Predict	10 (18) [69]	23 (28) [253]	137 (177)	0.774	59 (83)	2 (33)	0.711 (0.940)

^* See method section for details.
^# Values within square bracket indicates total number of proteins vaccine candidates (PVCs) predicted by respective software/server. Value within parenthesis indicates experimentally known protective vaccine candidates in that organism whereas the values in bold give the number of PVCs predicted by respective software/server from experimentally known protective vaccine candidates in that organism.
^$ Values indicate the number of PVCs predicted by respective software/server from 177 bacterial protective antigens (PAs) taken for evaluation.
^% Values indicate the number of PVCs predicted by respective software/server from 83 and 33 proteins for positive and negative datasets, respectively.

ISSN: 1471-2105