## Info

1, 3

1, 3

2, 4

2, 4

1, 3

Fractional Factorial Experimental Design Used for Screening of Growth Conditions and H2O2 Production by Vaginal Lactobacilli3

aBold numbers correspond to the central points of the experimental design. 1-4, days or block in which each condition was tested.

Table 2

Fractional Factorial Experimental Design Used for Screening of Growth Conditions and H2O2 Production by Vaginal Lactobacilli3

L. paracasei CRL 1289 L. crispatus CRL 1266

LAPTg MRS LAPTg MRS

Temp (°C) pH Without With Without With Without With Without With

37 6.5 1, 4, 5, 8 3, 2, 6, 8 2, 4, 6, 7 2, 3, 5, 7 2, 4, 5, 8 1, 3, 6, 7 1, 4, 6, 8 1, 3, 6, 8

aBold numbers correspond to the central points of the experimental design. The others represent the vertex of the experimental design. Without and With, without and with agitation. 1-8, days or block in which each condition was tested.

2. At time t, the reparametrized Gompertz model is expressed by the following function: Dt = D0 + Aexp{-exp [(^ x e/A) (k - t) + 1]} (1)

where Dt is the OD at time t, t is the time of growth in hours, D0 is the OD at t = 0, A is the difference between the initial and final ODs, ^ is the maximum specific growth rate (h), X is the duration time of lag phase in hours, and e is the base of the neperian logarithm (2.718281828).

### 3.3.2. Method of Growth Curve Estimation

1. Ordinary least squares: STATISTICA 5.0 software; the module used is Nonlinear Regression (see Note 15). For the algorithm of minimization of the loss function, the Rosenbrock/ quasi-Newton algorithm is used (see Note 16).

2. Bootstrap method (software S-PLUS): see basis explanation in Note 17. To calculate the standard errors of the parameters estimations, the bootstrapping technique is applied using repeated samples obtained from the original data, by sampling with replacement (17,18). The number of bootstrap samples was chosen as 1000 or 100 for each set of data of OD and CFU/mL obtained from the growth curve, respectively. SPSS 10 for Windows is used to apply all the calculations, and S-PLUS 200 is used to formulate the corresponding graphics (see Note 17).

3.3.3. Analysis of Variance (ANOVA)

To determine the statistically significant differences between the growth curves (log CFU/mL) or between the kinetics of antagonistic substances production under different culture conditions, the method of the analysis of variance (ANOVA) with repeated measurements was performed by using the statistical software MINITAB (see Note 18).