Data set of in silico simulation for the production of clavulanic acid and cephamycin C by Streptomyces clavuligerus using a genome scale metabolic model

Streptomyces clavuligerus (S. clavuligerus) is a Gram-positive bacterium which produced clavulanic acid (CA) and cephamycin C (CephC). In this data article, a curated genome scale metabolic model of S. clavuligerus is presented. A total of eighteen objective functions were evaluated for a better representation of CA and CephC production by S. clavuligerus. The different objective functions were evaluated varying the weighting factors of CA and CephC between 0, 1 y 2, whereas for the case of biomass the weight factor was varied between 1 and 2. A robustness analysis, by mean of flux balance analysis, showed five different metabolic phenotypes of S. clavuligerus as a function of oxygen uptake: (I) and (II) biomass production, (III) biomass and CephC production, (IV) simultaneous production of biomass, CA and CephC and (V) production of biomass and CA. Data of shadow prices and reduced cost are also presented.


Data
A total of twenty-four reactions were added for a better representation of the production of clavulanic acid (CA) and cephamycin C (CephC) by Streptomyces clavuligerus (see Table 1).
An array of eighteen combinations of different objectives functions varying the weighting factor of the slack variables was evaluated (see Table 2). The objective function was the maximization of biomass, CA and CephC. In order to evaluated the functionally of the objective functions the weighting factor of biomass, CA and CephC were varied (see experimental design). Table 2 also shows the metabolic scenarios where CA and CephC are produced or not.
The objective function No. 6 was the only one that included a metabolic phenotype that produced CA and CephC, simultaneously. Table 3 shows the fluxes of biomass, CA and CephC under different oxygen uptake for all eighteen combinations of the objective function (see also supplementary material 1). Fig. 1 shows five different metabolic phenotypes of S. clavuligerus as a function of oxygen uptake: (I) and (II) biomass production, (III) biomass and CephC production, (IV) simultaneous production of biomass, CA and CephC and (V) production of biomass and CA. See also supplementary material 2.
Specifications Table   Subject area Modelling and Simulation, Biotechnology More specific subject area

Flux balance analysis
Type of data Software COBRA Toolbox v3.0 running in a Matlab® environment, using Gurobi optimization software.

Data format
Simulated Experimental factors A total of eighteen objective functions were evaluated varying the weighting factors of CA and cephamycin C between 0, 1 y 2, whereas for the case of biomass the weight factor was varied between 1 and 2.

Experimental features
Diverse metabolic phenotypes for the production of CA and cephamycin C by Streptomyces clavuligerus, through a linear combination of the weighting factor on the objective function, were evaluated. Data source location Universidad del Valle, Escuela de Ingeniería Química, A.A. 25360 Cali, Colombia.

Data accessibility
Data is presented in this article only. Value of the data An updated genome scale metabolic model of Streptomyces clavuligerus is presented. The data will be useful for the understanding the metabolic phenotypes during the simultaneous production of clavulanic acid and cephamycin C by Streptomyces clavuligerus. This data will be useful to the researchers and scientific community working on clavulanic acid and cephamycin C production.

Flux balance analysis
Flux balance analysis (FBA) was used to determine metabolic states [12,13]. Loop law constrains was applied to all FBA simulation ensuring that infeasible loops ware not allowed [14]. The production of biomass, CA and CephC was used as objective function.

Optimization problem statement
Metabolic fluxes were quantified by means of a two-stage optimization approach, which is a combination of the maximization of the objective function and minimization of the overall flux [10,15,16]. The mathematical problem can be represented as follows: Stage one where Z is the objective function, S is the stiociometric matrix and v is the flux vector. w biomass , w CA and w Cephc are the weighting factors for biomass, intracellular flux of CA and CephC, respectively. v biomass , v CA intracellular and v Cephc intracellular are the biomass flux, intracellular flux of CA and CephC, respectively. v optbiomass , v optCA extracellular and v optCephc extracellular are the optimal values for biomass and extracellular flux of CA and CephC, respectively, that resulted from solving the problem stated at stage one. The first stage optimization problem was solved using a Gurobi solver, with a feasibility tolerance of 10 À6 , while the second stage was solved using the MATLAB's built-in fmincon solver, with a first order optimality and a maximum constraint violation within 10 À6 .
Different objective functions were evaluated varying the weighting factors of CA and CephC between 0, 1 y 2, whereas for the case of biomass the weight factor was varied between 1 and 2 (see Table 2).

Robustness analysis
A robustness analysis was carried out to evaluate the functionally of the objective function when the optimal flux of oxygen was varied [12,13]. The identification of possible gene knockout was made by sensitivity analysis using the concept of reduced costs. The reduced cost values represent the variation of the objective functions with respect to the fluxes related to each reaction and they are represented according to the equation (3). Additionally, the shadow prices were determined following the equation (4) [13,17].
Where, r i is the reduced cost, Z 0 is the optimal solution, v i is an internal flux that is not in the basis solution, p i is the shadow prices and b i is the exchange fluxes.

Computational tools
COBRA Toolbox v.3.0 synchronized with Matlab ® as programing environment, and the Gurobi optimizer 7.5.2 was used to solve all optimization problems [18].

Transparency document
Transparency document associated with this article can be found in the online version at https:// doi.org/10.1016/j.dib.2019.103992.