Table 1. Mann-Whitney U statistical test applied to performance measures of models generated from Normal training datasets.

Datasets		PPI	NO-PPI
		p-value	p-value
C×C′	Precision	1.15E-04	2.43E-06
C×C′	Recall	3.51E-06	0.002
C×C′	AUC	1.73E-06	1.73E-06
C×F+C	Precision	1.18E-11	1.61E-10
C×F+C	Recall	5.36E-10	8.36E-12
C×F+C	AUC	2.26E-11	2.26E-11
C×F′	Precision	8.41E-12	5.61E-10
C×F′	Recall	4.38E-09	0.000
C×F′	AUC	1.51E-09	1.51E-09
C×F′+C′	Precision	8.36E-12	1.61E-10
C×F′+C′	Recall	9.90E-10	7.21E-12
C×F′+C′	AUC	1.17E-10	1.17E-10
C′×F+C	Precision	7.54E-12	2.17E-11
C′×F+C	Recall	5.08E-11	0.000
C′×F+C	AUC	7.46E-12	7.46E-12
C′×F′+C′	Precision	7.49E-12	1.49E-11
C′×F′+C′	Recall	2.76E-11	7.56E-12
C′×F′+C′	AUC	8.33E-12	8.33E-12
F+C×F′+C′	Precision	9.00E-04	1.76E-02
F+C×F′+C′	Recall	*9.15E-02	0.002
F+C×F′+C′	AUC	2.46E-08	2.46E-08
F×C	Precision	1.72E-11	2.11E-10
F×C	Recall	7.28E-10	1.11E-11
F×C	AUC	1.40E-10	1.40E-10
F×C′	Precision	7.49E-12	2.18E-11
F×C′	Recall	3.84E-11	0.000
F×C′	AUC	4.59E-11	4.59E-11
F×F+C	Precision	*2.49E-01	*8.91E-02
F×F+C	Recall	*1.27E-01	*2.79E-01
F×F+C	AUC	*1.34E-01	*1.34E-01
F×F′	Precision	6.91E-04	1.61E-02
F×F′	Recall	*7.23E-02	0.001
F×F′	AUC	4.62E-08	4.62E-08
F×F′+C′	Precision	1.25E-02	*1.97E-01
F×F′+C′	Recall	*4.14E-01	1.61E-02
F×F′+C′	AUC	4.67E-06	4.67E-06
F′×C′	Precision	7.54E-12	1.11E-10
F′×C′	Recall	1.40E-10	0.000
F′×C′	AUC	1.16E-10	1.16E-10
F′×F+C	Precision	3.14E-05	2.31E-04
F′×F+C	Recall	2.86E-03	1.50E-04
F′×F+C	AUC	4.57E-10	4.57E-10
F′×F′+C′	Precision	*9.50E-02	*9.49E-02
F′×F′+C′	Recall	*9.75E-02	*0.162
F′×F′+C′	AUC	1.12E-02	1.12E-02

^*, no significant difference.