Table 3. Comparison of chemical shifts predicted using various methods to experimental values measured for Ubiquitin and GB3 and corrected for random coil effects.

The RMSD values are computed after linear regression. The predictions were done using CHARMM22/CMAP optimized structures using the GB/SA solvation model except for the first two rows (marked with ^a) where PM6-D3H+ optimized structures using the CPCM solvation model were used.

	Cα	Cβ	C′	Hα	HN	N
	RMSD (r)	RMSD (r)	RMSD (r)	RMSD (r)	RMSD (r)	RMSD (r)
Ubiquitin
DFTa	2.1 (0.62)	2.8 (0.56)	1.8 (0.85)	0.4 (0.83)	0.6 (0.81)	4.0 (0.80)
ProCS15a	2.0 (0.61)	2.2 (0.52)	1.7 (0.88)	0.4 (0.86)	0.6 (0.73)	4.4 (0.85)
ProCS15	1.7 (0.70)	2.0 (0.50)	1.7 (0.81)	0.4 (0.77)	0.6 (0.72)	4.0 (0.79)
CheShift-2	1.7 (0.59)	1.6 (0.62)
CamShift	1.1 (0.85)	1.3 (0.71)	1.0 (0.81)	0.3 (0.73)	0.5 (0.69)	3.0 (0.63)
PPM_One	0.7 (0.93)	1.1 (0.80)	0.9 (0.87)	0.2 (0.88)	0.4 (0.73)	2.2 (0.81)
Sparta+	0.7 (0.93)	1.1 (0.82)	0.8 (0.88)	0.2 (0.86)	0.4 (0.72)	2.2 (0.81)
shAIC	0.7 (0.94)	1.1 (0.82)	0.8 (0.89)	0.3 (0.83)	0.5 (0.71)	2.3 (0.79)
ShiftX2	0.5 (0.97)	0.7 (0.91)	0.5 (0.96)	0.1 (0.97)	0.3 (0.91)	1.8 (0.88)
GB3
DFTa	2.1 (0.71)	2.4 (0.53)		0.4 (0.76)	0.6 (0.86)	4.6 (0.78)
ProCS15a	1.8 (0.73)	2.1 (0.42)		0.4 (0.75)	0.7 (0.85)	4.8 (0.88)
ProCS15	1.6 (0.70)	2.0 (0.42)		0.3 (0.85)	0.6 (0.76)	4.3 (0.86)
CheShift-2	1.7 (0.68)	1.8 (0.53)
Camshift	1.2 (0.81)	1.0 (0.83)		0.3 (0.85)	0.4 (0.82)	3.3 (0.54)
PPM_One	1.0 (0.87)	0.9 (0.87)		0.3 (0.91)	0.4 (0.89)	2.3 (0.79)
Sparta+	1.0 (0.87)	1.0 (0.86)		0.3 (0.89)	0.4 (0.88)	2.8 (0.70)
shAIC	1.0 (0.88)	1.0 (0.85)		0.3 (0.87)	0.4 (0.83)	2.3 (0.79)
ShiftX2	0.6 (0.96)	0.7 (0.93)		0.1 (0.97)	0.1 (0.98)	2.3 (0.79)