Table 3 shows the statistical analyses carried out using Statistica 9 software. All the variables shown in the regression Dinaciclib concentration coefficients table were significant as shown by their p-values (lower than the α significance level), the linear sardine-FFA:SO mole ratio showing the greatest effect on the response of all the variables studied. The negative effect of this variable on the response means that a high proportion of FAs as compared to the soybean oil (sardine-FFA:SO mole ratio of 3:1) allowed for a more effective incorporation of the EPA and DHA. The initial water content of the enzyme showed a positive effect on the response, which means that the higher value

of initial water content of the enzyme tested (0.86) allowed for higher EPA + DHA incorporation levels. Although the reaction time did not directly influence the reaction rate (non-significant variable), it became significant when the initial water content of the enzyme varied, given

that the interaction between the reaction time and initial water content of the enzyme was statistically significant. Fig. 1 shows the contour www.selleckchem.com/products/Trichostatin-A.html diagram for the incorporation of EPA + DHA during the acidolysis reaction. The calculated t values for the variables which showed positive effects (t = 11.93 for the initial water content of the enzyme and t = 6.27 for the reaction time versus the initial water content of the enzyme) were higher than the tabulated t values (2.35), showing that the effects of these variables and/or interactions were greater than the standard error, and thus significant. In the same way, the calculated t values for the variables and interactions which showed negative effects (t = −19.76 for the substrate mole ratio and t = −9.24 for the substrate mole ratio versus the initial water content of the enzyme) were significant, as they were lower than the

negative tabulated t value (−2.35). The calculated value for F (15.87) was significant and the percent of variation explained by the model (R2) was highly satisfactory, so it was concluded that the experimental model, expressed by the equation z = 2.08 + 0.38x − 8.61y − 0.05 × 24x + 0.37 × 24y + 4.90xy − 1.36, adequately fitted the Non-specific serine/threonine protein kinase experimental data. The deviations between the values predicted by the experimental model and the experimental values observed in the acidolysis reactions were normally distributed, i.e., positive and negative deviations around the predicted values appear in the same proportion, showing a non-tendentious behaviour in their distribution. The observed values were very close to the predicted values, meaning that the experimental model was adequately fitted. Fig. 2 shows the profiles of the TAGs in the soybean oil before and after acidolysis as obtained by EASI(+)-MS. The characteristic soybean oil TAG profiles were detected mainly as [TAG + Na]+ ions (Fig.