Computing gene knockout strategies (MCS) for different degrees of growth-coupled production

Supplemetary code for:

Systematizing different notions of growth-coupled product synthesis and a single framework for computing corresponding strain designs

Philipp Schneider, Radhakrishnan Mahadevan, Steffen Klamt

2021/03/16

corresponding author: [email protected]

Download code

Download latest CellNetAnalyzer

Added Features (CellNetAnalyzer):

1. Definition of optimality constraints (r _i = r _i,max) for describing desired and undesired behavior (in addition to static linear constraints of the form T r ≤ t and D r ≤ d). Such constraints can be used to define MCS setups for potentially and weakly growth-coupled product synthesis. For potentially growth-coupled production, one defines the desired flux states to be growth-optimal and compliant with minima for the attainable growth rate and product synthesis rate. For weakly growth-coupled product synthesis, one uses a target system in which growth-optimality is combined with the static constraint r_P = 0. MCS will seek to eliminate flux states in which maximum growth and zero production coincide.

Software Requirements:

1. MATLAB 2016b® or later

2. IBM ILOG® CPLEX® 12.7, 12.8, 12.9 or 12.10 (Make sure to use compatible CPLEX® and MATLAB® versions. Version 12.10 is recommended. CPLEX® 20.1 DOES NOT WORK PROVIDE A MATLAB API AND CANNOT BE USED HERE)

3. CellNetAnalyzer 2021.1 or later

4. Set up CellNetAnalyzer to access the CPLEX-Matlab-API (as described by CellNetAnalyzer manual)

Getting Started:

1. Download this project to your computer (see release page) and extract all files.
2. Download the E. coli model iML1515 from the Bigg-database. Place it in the project folder.
3. Start MATLAB.
4. Add the main directory of your installation of the CellNetAnalyzer toolbox to your MATLAB path.
5. Add the project folder and the subdirectory 'functions' to your MATLAB path.

Script Files:

1. MCS_1_coupling_degrees.m

   Computes exhaustively all minimal cut sets up to three gene knockouts for the potentially, weakly and directionally growth-coupled and substrate-uptake coupled production of ethanol with E. coli. A subnetwork/core network (ca. 600 reactions) of the iML1515 is used to shorten the computation runtime. Finally, the relationship between the MCS sets of the different coupling types is plotted. The user can set/unset the flag to de-/activate the minimum ATP maintenance demand.

2. MCS_2_smallest.m

   Computes the smallest minimal cut set for the potentially, weakly and directionally growth-coupled and substrate-uptake coupled synthesis of 12 different products with E. coli. The genome-scale model iML1515 is used. For heterologous products, the pathways are added automatically. As the runtime directionally depends on the random seed used in the MCS computation, multiple computations are run with a time limit. To avoid memory problems, the computation is run in a seperate MATLAB instance. By default, the script runs with the following settings: (1) ethanol production (2) weak growth-coupled production (3) 6 iterations (4) 4 hours per computation (5) run each computation in a new process each (6) maximum of 60 knockouts (7) return after finding 1 solution. The user can change these settings to compute MCS for different products or coupling strengths.

3. MCS_3_any.m

   Computes a random minimal cut set for the potentially, weakly and directionally growth-coupled and substrate-uptake coupled synthesis of 12 different products with E. coli. The genome-scale model iML1515 is used. For heterologous products, the pathways are added automatically. As the runtime directionally depends on the random seed used in the MCS computation, multiple computations are run with a time limit. To avoid memory problems, the computation is run in a seperate MATLAB instance. By default, the script runs with the following settings: (1) ethanol production (2) weak growth-coupled production (3) 12 iterations (4) 2 hours per computation (5) run each computation in a new process each (6) maximum of 60 knockouts (7) return after finding 1 solution. The user can change these settings to compute MCS for different products or coupling strengths.

4. MCS_4_ACP.m

   Computes exhaustively all minimal cut sets up to three gene knockouts for the directionally growth-coupled, ATP-coupled and substrate-uptake coupled production of ethanol with E. coli. A subnetwork/core network (ca. 600 reactions) of the iML1515 is used to shorten the computation runtime. The minimum ATP maintenace demand from the original iML1515 model is omitted. Finally, the relationships between the MCS sets of the different coupling types are plotted.

Auxiliary functions:

5. functions/load_pathway.m Contains the pathways for all heterologous products. This function returns the species and reactions that need to be added to the iML1515.

6. functions/block_non_standard_products.m Shut metabolite exchanges for all atypical products of E. coli.

7. functions/check_mass_balance.m

8. functions/compare_mcs_sets.m Between two sets of MCS, identify matching MCS, MCS subsets and supersets.

9. functions/compute_MCS_ext MCS computation function that is run by a new MATLAB instance.

10. functions/plot_mcs_relationships.m

Model files:

11. core.mat - Indicates species and reactions within the iML1515 that are part of E. coli's core metabolism

12. iML1515geneNames.mat - contains a map of gene names and b-numbers. This helps to generate a user friendly output.

13. iML1515.mat - Required for computation, but not provided in this repository. Please download from the Bigg-database

Relevant new (API) functions included in the most recent release (2021.1) of the CellNetAnalyzer toolbox :

• CNAgeneMCSEnumerator3

  Function wrapper for CNAMCSEnumerator3 that allows the computation of gene-MCS with all features introduced in CNAgeneMCSEnumerator2 and additionally the option to define optimality constraints.

• CNAMCSEnumerator3

  Features all functions of CNAMCSEnumerator2 and additionally the option to define optimality constraints to describe desired or undesired flux vectors.

Remarks:

• As their runtimes are usually very long, scripts MCS_2_smallest.m and MCS_3_any.m compute only one specific setup (by default: ethanol, wGCP). The computation for other products or coupling strengths must be defined by the user. The according lines are marked in the comments of the scripts.