Table 1.

Proteins used in this study

	Name	Abbrev.	PDB code	# of SSEs
(a) The testing set
1	Apo-myoglobin	apoMb	1mbo	8
2	Barnase	Bnase	1a2p	8
3	Cytochrome c	Cytc	1hrc	5
4	T4 lysozyme	T4lzm	2lzm	14
5	Ribonuclease T1	RnaseT1	9rnt	8
6	α-Lactalbumin	ha-LA	1hml	12
7	Chymotrypsin inhibitor 2	CI2	2ci2	5
8	Ubiquitin	Ubq	1ubi	7
9	Bovine pancreatic trypsin inhibitor	BPTI	5pti	5
10	Interleukin-1β	IL-1b	1i1b	14
11	Hen egg-white lysozyme	HEWL	1hel	10
12	Equine lysozyme	Eqlzm	2eql	10
13	Protein A, B-domain	pAB	1bdd	3
14	Ribonuclease A	RnaseA	1rbx	10
15	Guinea pig α-lactalbumin	gpa-LA	1hfx	9
16	B1 immunoglobulin-binding domain protein G	GB1	1pga	5
17	B1 immunoglobulin-binding domain protein L	LB1	2ptl	5
18	Cardiotoxin analog III	CTX-3	2crt	5
19	Tendamistat	Tnds	2ait	6
20	Single chain antibody fragmenta	scFv	2mcp	23
21	Human acidic fibroblast growth factor-1	hFGF-1	2afg	14
22	Cytochrome c551	pacc551	351c	5
23	Outer surface protein A	ospA	1osp:O	24
24	Ovomucoid third domain	OMTKY3	1iy5	4
25	Chicken src SH3 domain	cSH3	1srm	6
26	CheY	CheY	3chy	10
27	Human carbonic anhydrase I	HCA-1	1hcb	22
(b) The training set
28	Staphylococcal nuclease	Snase	1stn	9
29	Ribonuclease H	RnaseH	2rn2	10

^aFor consistency with previous studies [13,15], we consider only the residues 1-115 in chain L and 1-122 in chain H.