Modular prediction of protein structural classes from sequences of twilight-zone identity with predicting sequences

Table 3 Number of features selected for each structural class for different categories of features.

Class	AA sequence	PSSM	PSSM and predicted secondary structure	Predicted secondary structure	Collocations of H and E segments	Total
α	8	26	52	21	10	117
β	2	28	9	8	6	53
α/β	0	0	17	17	12	46
α+β	2	11	101	27	22	163
Multi-domain	3	17	43	26	16	105
Membrane	6	16	18	6	0	46
Small	6	6	3	3	0	18

ISSN: 1471-2105