Table 2.

The regression analysis of a full second-order polynomial model for the optimized titer of D-PA by the E. coli W3110.

Source	Sum of Squares	df	Mean Square	F Value	P value
Model	27.16	9	3.02	18.97	0.0004
X1 (glucose)	9.89	1	9.89	62.19	< 0.001**
X2 (β-alanine)	4.67	1	4.67	29.34	0.001**
X3 ((NH4)2SO4)	0.06	1	0.06	0.4	0.5451
X1X2	1.38	1	1.38	8.66	0.0216*
X1X3	0.03	1	0.03	0.19	0.6781
X2X3	0.01	1	0.01	0.05	0.8318
X1^2	3.26	1	3.26	20.48	0.0027*
X2^2	6.08	1	6.08	38.26	0.0005*
X3^2	0.8	1	0.8	5.01	0.0603
Residual	1.11	7	0.16
Lack of Fit	0.81	3	0.27	3.61	0.1236
Pure Error	0.3	4	0.08
Cor Total	28.27	16

^aR² = 0.9836, adj-R² = 0.9626, pred-R² = 0.9359, Adequate precision = 12.265, CV = 1.25%,

*significant (p-value < 0.05)

** highly significant (p-value < 0.01)