Table 1.

Comparison of the leave-one-out cross-validations using the DichroIDP method with the dictionary of secondary structure of proteins (DSSP)²³ assignment method.

a) IDP175				b) IDP175t
	r	δ	ζ		r	δ	ζ
H	0.9270	0.0801	2.6633	H	0.9214	0.0831	2.5668
E	0.8543	0.0886	1.9186	E	0.8422	0.0920	1.8476
T	0.5342	0.0613	1.1365	T	0.5411	0.0599	1.1631
D	0.9322	0.0649	2.6987	D	0.9364	0.0617	2.8400
c) SP175				d) SP175t
	r	δ	ζ		r	δ	ζ
H	0.9299	0.0771	2.7191	H	0.9233	0.0807	2.5977
E	0.8398	0.0871	1.8354	E	0.8018	0.0956	1.6715
T	0.3691	0.0543	1.0353	T	0.3915	0.0535	1.0510
D	0.5945	0.0535	1.3051	D	0.7282	0.0484	1.4414
e) CDPro42				f)	SP175+17 (from BeStSel12)
	r	δ	ζ		r	δ	ζ
H	0.9175	0.0870	2.5001	H	0.9087	0.0897	2.3924
E	0.6980	0.1145	1.3710	E	0.8016	0.1042	1.6712
T	0.5296	0.0771	1.0873	T	0.4454	0.0531	1.0967
D	0.7385	0.1550	1.3511	D	0.6623	0.054	1.3077

The following reference datasets were used: a) IDP175; b) IDP175t (low wavelength cut off 190 nm); c) SP175⁸, d) SP175t (low wavelength cut off 190 nm); e) CDPro42¹⁰ and f) SP175+¹⁷. The statistical parameters reported are: r, the Pearson’s correlation coefficient; δ, the root mean squared deviation, and ζ the ratio of δ over the population standard deviation as defined in the main text. The cross-validation values for all of the reference datasets/assignment methods are similar for helix and sheet secondary structures, but the disordered structure contents are very much improved using the IDP175 and IDP175t reference datasets. H,E,T, and D refer to the helical, sheet, turn, and disordered components, respectively.