We wanted to use elementary flux mode analysis to identify targets for gene knockout/gene overexpression. Elementary flux mode analysis allows us to calculate the elementary flux modes, which are analogous to the 'metabolic freedom' of the metabolic model. Using the EFMs, we are able to understand all of the "potential capabilities" of the organism. We then examine the metabolic network for different things we want to have done, and things we want to eliminate in order to increase pABA production. Using this, we select targets for knockout/overexpression.
In order to calculate the EFMs, we used efmtool [1] from ETH Zurich. We also used several MATLAB scripts from Dr. Nils Averesch to calculate product yields from the EFM model.
We were interested in finding ways to produce a larger amount of pABA. In a standard, unmodified metabolic network, the amount of pABA being produced is relatively little. The calculated product yield was:
Product_max_without_RX = 49.6064
The biomass vs. product yield graph is shown below:

We generated a metabolic network where R34 (Phosphotransferase system (EC 2.7.1.69)) was removed. Our reasoning was that glucose is preferentially taken up through the phosphotransferase system, which consumes PEP. In our case, PEP is also the precursor to the Shikimate pathway. By changing the glucose uptake mechanism, we expect to see an increase in the amount of pABA. With R34 removed, the results were:
Product_max_without_RX = 92.4829