Homology modelling of protein-protein complexes: a simple method and its possibilities and limitations

BMC Bioinformatics

Table 2 Discriminating experimental complexes from threaded models

Superfamily identifiers	Superfamily names	sequence Id ranges (%)	number of complexes	number of tests	number of successes^a
					DFIRE_β	Launay
a.133.1	Phospholipase A2	33–99	11	110	77	62
d.17.4	NTF2-like	11–98	13	156	140	92
b.29.1	lectins	33–99	15	210	205	148
c.61.1	Ribosyltransferase	15–99	10	90	80	65
e.3.1	β lactamase-like	9–100	11	110	107	59
d.5.1	RNase A-like	25–94	6	30	26	17
c.2.1	Rossmann-fold	11–43	14	182	174	152
a.118.6 a.102.4	Prenylyltransferase	94–99	8	56	56	27
	RhoGDI/Ras	54–94/73–100	4	12	11	6
	FGF receptor/FGF	20–35/22–62	3	6	6	6
	Trypsin/inhibitor	30–100/10–100	12	132	127	87
	Trypsin/inhibitor	65–83/56–71	3	6	4	1
	Peptidase M10/TIMP	42–98/50–100	3	6	4	6
	Trypsin/inhibitor	14/98	2	2	2	1
	Trypsin/inhibitor	39/42	2	2	2	0
	Collicin, Pyocin/HNH	39/42	2	2	2	1
	Elongation factor/EF-TS	20–57/6–22	4	12	12	7
		Total	123	1124	1035	737

^aThe number of tests giving a positive S_threadscore. Both the 6-class, residue-based and the DFIRE_β, atom-based were used for scoring.

ISSN: 1471-2105