Table 1.

Performance comparison of server groups participated in CASP13 refinement category

Group names	Mean improvement of Model 1 / Best among Model 1–5a
	GDT-HA	GDC-SC	LDDTb	– MolProbity
Seok-server (GalaxyRefine2)	+1.46 / +2.68	+3.45 / +5.06	+2.55 / +3.21	+1.47 / +1.59
Bhattacharya-Server (29)	–0.37 / +1.75	+1.70 / +3.55	+0.86 / +1.79	+1.19 / +1.34
YASARAc	–1.21 / –1.21	+1.69 / +1.69	+0.57 / +0.57	+1.60 / +1.60
MUFold_server	–2.28 / –1.54	–0.69 / +1.17	–0.63 / –0.26	–0.40 / –0.17
3DCNN	–11.44 / –3.28	–6.52 / –1.05	–6.83 / –3.43	+0.65 / +0.87

^aAll evaluation values were obtained from CASP official homepage, http://predictioncenter.org/casp13/results.cgi.

^bLDDT values were re-scaled to the range of [0, 100].

^cYASARA group submitted only one model per target.