Clarification on the flexibilizeEnzConcs Function and Its Impact on Enzyme Concentrations and kcat Values

[HossFir]

As I understand the flexibilizeEnzConcs function, it increases the concentration of bottleneck enzymes to make the objective function feasible. When proteomics data is integrated, the function increases the upper bound of these bottleneck enzymes. My question is: does this flexibilized enzyme concentration replace the original measured proteomics data as the new upper bound, or is it just an arbitrary adjustment to make the model feasible? It was mentioned that the values in ecModel.ec.concs will not be affected, so after reporting the relaxed enzyme concentrations, is the enzyme usage (such as topAbsUsage) or total protein usage calculated based on the non-relaxed enzyme concentrations obtained from the proteomics data (as the relaxed enzyme concentration is higher than proteomics)?

Additionally, after running flexibilizeEnzConcs, some bottleneck kcats are identified. We then increase these kcats via the CustomKcats.tsv . Do we do this because relaxing the enzyme concentrations of the bottleneck kcats does not work (as we increase the enzyme concentration repeatedly), so we increase the corresponding kcats to make the model feasible?