PubChem3D: conformer ensemble accuracy

Table 2 Linear behavior of average conformer model accuracy as a function of RMSD ^cluster value

Accuracy measure	Data series	a	b	σ _a	σ _b	σ _y	R ²
RMSD	Before	−0.06	0.63	0.087	0.061	0.11	0.93
	After	−0.07	0.90	0.083	0.059	0.11	0.97
	Difference	−0.01	0.27	0.045	0.032	0.06	0.90
ST ^ST-opt	Before	1.06	−0.15	0.020	0.014	0.03	0.94
	After	1.06	−0.21	0.025	0.018	0.03	0.95
	Difference	−0.00	0.06	0.012	0.009	0.02	0.85
ComboT ^ST-opt	Before	2.13	−0.53	0.057	0.040	0.07	0.96
	After	1.99	−0.59	0.088	0.062	0.11	0.92
	Difference	0.14	0.06	0.060	0.043	0.08	0.19
CT ^CT-opt	Before	1.10	−0.34	0.038	0.027	0.05	0.95
	After	1.01	−0.36	0.050	0.035	0.06	0.93
	Difference	0.09	0.02	0.036	0.025	0.05	0.07
ComboT ^CT-opt	Before	2.16	−0.54	0.058	0.041	0.07	0.96
	After	2.03	−0.60	0.091	0.064	0.12	0.92
	Difference	0.13	0.06	0.065	0.046	0.08	0.17

Results of linear least-squares fitting of the average conformer model accuracies vs. RMSD clustering threshold (RMSD^cluster) to the form of “y = a + bx”. The sigma values (σ_a, σ_b, and σ_y) correspond to the standard deviation of the fit to the predicted “a”, “b”, and “y” values. The “x” values are the discrete RMSD^cluster values and the “y” values are the corresponding average accuracy measures found in the data series from panel (d) in Figures 8, 9, 10, 11, and 12.

ISSN: 1758-2946