BMC Bioinformatics

Table 3 The top 20 features selected by correlation coefficient (CC-) and information entropy (IE-) based methods

From: Prediction of mucin-type O-glycosylation sites in mammalian proteins using the composition of k-spaced amino acid pairs

Site	S		T
Top 20 features	CC	IE	CC	IE
1	ST	PXXS	TXXP ^a,b	TXXP
2	SXXP	PXP	PT	PT
3	PXXXXS	PS	AXXXP	TT
4	PXXS	SXS	PXT	TXXXT
5	TXXP	SXXP	PP	TXXXXT
6	PXP	ST	TP	TXT
7	PXXXXP	SXXXP	AXP	PXT
8	SXXXP	PXXXP	PXXXXP	TP
9	SXP	SXXXS	TXXXXP	PP
10	TXXXP	PXXXXP	TXA	PXP
11	PS	PXXXXS	PXXXP	TS
12	SXXXT	SXXXXP	SXXXP	TXXXXP
13	TP	TS	SXXT	PXXXXP
14	TXXXXS	SS	TT	SXXT
15	TS	SP	PXXA	PXXXP
16	TXXA	SXP	SA	SXXXXT
17	PXT	PXT	PXA	TXXS
18	PXXXP	TXXXXS	TXXXP	ST
19	PP	TXXP	PXP	TXXXP
20	PXXP	PXXP	AXXXXT	PXXXT

^a TXXP represents a 2-spaced amino acid pair of TP, where X stands for any amino acid. The same representation was applied to other k-spaced amino acid pairs. ^b The k-spaced amino acid pairs in bold type mean they are consistently ranked as the top 20 features by both feature selection methods.

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com