Table 1.

The major type of features and the number of selected contributing features in each referenced study

Feature category	Feature description	Feature dimentionality	Sources	Number of selected contributing features in each referenced study
Sequence properties	Sequence length	1	Uniprot [256], Profeat [257]	–
	Amino acid composition	20		Hong et al. (2010):13 [22]; Wang et al. (2016):4 [23]; Wang et al. (2013):5 [25]; Sun et al. (2015):3 [24]
	Di-peptides composition	400		Hong et al. (2010):12 [22]; Wang et al. (2016):33 [23]; Wang et al. (2013):25 [25]; Sun et al. (2015):20 [24]
	Normalized Moreau–Broto autocorrelation	240		Wang et al. (2016):8 [23]; Wang et al. (2013):4 [25]; Sun et al. (2015):5 [24]
	Moran autocorrelation	240		Wang et al. (2016):8[23]; Wang et al. (2013):5[25]; Sun et al. (2015):7 [24]
	Geary autocorrelation	240		Wang et al. (2016):8 [23]; Wang et al. (2013):6 [25]; Sun et al. (2015):6 [24]
	Sequence order	160		Wang et al. (2016):10 [23]; Sun et al. (2015):6 [24]
	Pseudo amino acid composition	50		Hong et al. (2010):11 [22]; Wang et al. (2016):3 [23]; Sun et al. (2015):3 [24]
Physicochemical properties	Hydrophobicity	21	Profeat [257], Fldbin [258], ExPASy Tools [259]	Cui et al. (2008):8 [20]; Hong et al. (2010):4 [22]; Wang et al. (2013):2 [25]; Sun et al. (2015):5 [24]
	Normalized Van der Waals volume	21		Cui et al. (2008):11 [20]; Hong et al. (2010):5 [22]; Wang et al. (2016):2 [23]
	Polarity	21		Cui et al. (2008):12 [20]; Hong et al. (2010):4 [22]; Wang et al. (2013):2 [25]; Sun et al. (2015):1 [24]
	Polarizability	21		Cui et al. (2008):8 [20]; Hong et al. (2010):4 [22]; Wang et al. (2016):1 [23]; Wang et al. (2013):1 [25]; Sun et al. (2015):1 [24]
	Charge	21		Cui et al. (2008):11 [20]; Hong et al. (2010):5 [22]; Wang et al. (2016):1 [23]; Wang et al. (2013):1 [25]; Sun et al. (2015):1 [24]
	Secondary structure	21		Cui et al. (2008):13 [20]; Hong et al. (2010):4 [22]; Wang et al. (2016):1 [23]; Wang et al. (2013):2 [25]; Sun et al. (2015):2 [24]
	Solvent accessibility	21		Cui et al. (2008):6 [20]; Hong et al. (2010):3 [22]; Wang et al. (2016):1 [23]; Sun et al. (2015):2 [24]
	Unfoldability	1		Cui et al. (2008):1 [20]; Hong et al. (2010):1 [22]
	Fldbin charge	1
	Hydrophobicity	1		Cui et al. (2008):1 [20]; Wang et al. (2013):1 [25]
	Longest disordered regions	1		Cui et al. (2008):1 [20]; Wang et al. (2016):1 [23]
	Isoelectric point	1		Hong et al. (2010):1 [22]; Wang et al. (2016):1 [23]; Sun et al. (2015):1 [24]
	Charge	1		Cui et al. (2008):1 [20]; Hong et al. (2010):1 [22]; Wang et al. (2016):1 [23]
	Molecular weight	1
	Percentage of disordered region	1		Hong et al. (2010):1 [22]
	Percentage of disordered residues	1
	Relative surface accessibility	3		Cui et al. (2008):3 [20]
Domains/motifs properties	Beta-barrel transmembrane (BBTM) score	1	SingalP [260], TMB-Hunt [261], TatP [262], Phobius [263], NetOglyc [264], NetNGlyc [265]	Cui et al. (2008):1 [20]
	Log P BBTM/Non-BBTM protein ratio	1		Cui et al. (2008):1 [20]
	Twin-arginine signal peptide	1		Cui et al. (2008):1 [20]
	Transmembrane domains	1		Cui et al. (2008):1 [20]; Wang et al. (2016):1 [23]; Sun et al. (2015):1 [24]; Hong et al. (2010):1 [22]
	Signal peptide	1		Cui et al. (2008):1 [20]; Hong et al. (2010):1 [22]; Sun et al. (2015):1 [24]
	Glycosylation number and presence	4		Cui et al. (2008):1 [20]; Hong et al. (2010):2 [22]
	C-mannosylated	1
	Phosphorylation sites	1
	Cleavage site	2		Wang et al. (2016):1 [23]
	Subcellular location	4
	Percentage of coil-content	1		Hong et al. (2010):1 [22]
Structural properties	Secondary structural content	4	SSCP [266], Radius of Gyration	Cui et al. (2008):1 [20]; Wang et al. (2016):1 [23]; Sun et al. (2015):2 [24]
	Radius gyration	1		Cui et al. (2008):1 [20]
	Radius	1		Wang et al. (2016):1 [23]; Wang et al. (2013):1 [25]; Sun et al. (2015):1 [24]