Table 1.
Comparison of frequently used approaches in fluxomics.
Parameter A is used in the extended Debye-Hückel equation.
Table 2.
Factors considered in mod-matTFA.
Values 0/1 refer to the binary codification for the full factorial design (S3 Table). In total, 26 combinations were tested.
Fig 1.
Calculation of the parameter A.
The red line refers to the temperature-dependent function (Eq 5), whereas the surface is the temperature/salinity-dependent function (Eq 6).
Table 3.
The type of analysis depended on the nature of the response: classification was selected for TFA (full), TFA (match 13C-MFA), concentration values (full) and concentration values (match experimental), and regression for r (fluxomics) and r (metabolomics). The former was suited for categorical responses (i.e. which solution is achieved, as shown in S4 Table), and the latter for continuous responses (for Pearson's r, from -1 to +1).
Table 4.
Tests with the highest score in the joint ranking.
The full list is available in (S4 Table). *(run #3) reflects the conditions used in the original matTFA.
Table 5.
Quantified metabolites in the top 50% of PageRank (PR) based analysis.
The last position in the ranking (#622) was L-Tyrosine (PR score = 0.0004), which had been quantified. The full list can be found in (S4 Dataset).
Fig 2.
Set of carboxylation/decarboxylation reactions including phosphoenolpyruvate, pyruvate, oxaloacetate, and malate. Arrows indicate the expected direction of carbon fluxes. Boxes refer to reactions: blue when they are defined in both the GSM and the metabolic network used for 13C-MFA, and orange when they are exclusively considered in the GSM. In the latter case no mapping was possible (S1 Table).
Table 6.
Flux pattern changes between 13C-MFA data and matTFA predictions.