Fig 1.
Schematic diagram of workflow carried out in the study.
Key steps (red) of the workflow for constructing CIs, with necessary inputs in purple. When the necessary tasks (green) are completed for each step (red), one can move to the next step, repeating the cycle when the confidence intervals do not pass. Only after robust CIs are constructed should one move on to the applications in blue to prevent incorrect conclusions.
Fig 2.
Control coefficients for glucose transport (GLCptspp) and their CIs derived using different methodologies.
The diamonds indicate the mean of the FCCs in decreasing order of absolute mean. CIs were derived using (A) univariate t-distributions for n = 1 degrees of freedom, (B) Bonferroni, (C) exact normal, and (D) bootstrapping approaches (for more information, see Materials and Methods). The lower and upper whiskers correspond to the CI range. The CIs are blue when they include zero at the 95% confidence level and red if they do not. Numerical values are given in the supplementary information (S2 Table).
Fig 3.
Top flux control coefficients of glucose uptake (GLCptspp) with confidence intervals determined using four different statistical approaches.
The top 10 FCCs based on absolute mean are reported with diamonds. The whiskers indicate the CIs for univariate (magenta), Bonferroni (blue), exact normal (red), and bootstrapping (black). The reader is referred to the Materials and Methods for technical details on CI computation. Numerical values are given in the supplementary information (S2 Table).
Fig 4.
Case study: Differences in means using bootstrapping.
Comparison of the differences in means of 15 FCCs for GLCptspp across four cases using the bootstrapping method (see Materials and Methods). The whiskers indicate the CIs, and the diamonds report the estimates of the differences in means. The CIs are blue with a hollow diamond when they include zero at the 95% confidence level and red with a plain diamond if they do not. The tests were carried out globally on the 90 estimates, even though we report each case comparison as a separate plot. Numerical values are given in the supplementary information (S3 Table).
Fig 5.
Case study: Differences of means using the univariate method.
Comparison of the differences in means of 15 FCCs for GLCptspp across four cases using the univariate method (see Materials and Methods). The whiskers indicate the CIs, and the diamonds report the estimates of the differences in means. The CIs are blue with a hollow diamond when they include zero at 95% confidence level and red with a plain diamond if they do not. The tests were carried out globally on the 90 estimates even though we report each case comparison as a separate plot. Numerical values are given in the supplementary information (S3 Table).