Table 2.

An overview of the protein prediction and application of disease biomarker. N/A = not application (no application discussion in this article)

Study	Body fluid	Algorithm	Size of training data set	# of selected features	Performance	Application outcome	Ref
Cui et al. (2008)	Bloodstream	SVM	6696 (151/6545)	85	0.94 (AUC)	13 biomarkers in gastric cancer; 26 biomarkers in lung cancer	[20]
Hong et al. (2011)	Urine		3940 (1313/2627)	74	0.90 (AUC)	Six biomarkers in gastric cancer	[22]
Wang et al. (2013)	Saliva		7077 (261/6816)	55	0.81 (AUC)	37 biomarkers in breast cancer	[25]
Sun et al. (2015)	Saliva		2757 (556/2201)	68	0.90 (Accuracy)	29 biomarkers in head and neck cancer	[24]
Wang et al. (2016)	Urine		2000 (1000/1000)	87	0.93 (AUC)	29 biomarkers in lung cancer	[23]
Liu et al. (2010)	Bloodstream	Ranking	11 394 (253/11 141)	85	0.66 (AUC)	N/A	[21]
Hu et al. (2011)	Body fluids	PPI	529	N/A	0.96 (Accuracy)	N/A	[26]