Biomedical Engineering Reference
In-Depth Information
Table 8. Continued
Natural values
TAA
(EU/gds)
Observed
response
M
IS
TS
Predicted value
84
6.5
40.0
296.9
318.0
72
4.3
40.0
245.0
249.8
72
8.8
40.0
384.4
383.8
72
6.5
9.5
386.1
361.0
72
6.5
70.5
487.8
513.8
72
6.5
40.0
451.1
437.4
72
6.5
40.0
428.7
437.4
72
6.5
40.0
455.6
437.4
72
6.5
40.0
437.6
437.4
72
6.5
40.0
424.2
437.4
72
6.5
40.0
426.8
437.4
Significance analysis of coefficients
Factor
Coefficient
t
-value
Constant
437.0
81.83*
M
47.5
12.71*
IS
43.9
11.74*
TS
50.1
13.39*
M·IS
6.8
1.44
M·TS
16.3
3.47*
IS·TS
-0.6
0.12
M
2
-82.5
20.39*
IS
2
-51.9
12.81*
TS
2
0.7
0.18
*
Significant at
t
(
α
< 0.05;
υ
= 5) > 2.57.
Taking into account the results obtained in this second optimization experiment, a pH
value of 5.9 and a temperature of 32.5ºC were used for further optimization studies.
The following study was focused on the determination of the effects of three factors (total
sugars concentration, inoculum size and moisture content) on amylase production in the
optimal conditions calculated before (particle size, pH and temperature) by using a new
second-order orthogonal design. Table 8 summarizes the experimental and predicted
responses along with significance of the model terms.
With the model coefficients acceptable according to the Student
t
test (
α
< 0.05), the
equation obtained is significant in Fisher´s
F
tests (
α
= 0.05) applied to the quotients total
error/experimental error and lack of fitting/experimental error (Table 9). The value of the
coefficient of determination (0.969) ensures that the model fits satisfactorily the data and
indicates that 96.9% of the variability in
TAA
production could be explained by the empirical
model (3):
TAA
(EU/gds) = 437.0 + 47.5
M
+ 43.9
IS
+ 50.1
TS
+ 16.3
M
·
TS
- 82.5
M
2
- 51.9
IS
2
(3)
