Description
- A customisable sequential gene scanner for optimising combinations of both gene deletions and upregulations using flux balance analysis
- Allows adapting the number of affected genes, the loop selection percentile, the upregulation factor & the growth rate threshold
- Created by: @LucoDevro, Lucas De Vrieze, Department of Chemical Engineering, KU Leuven, Heverlee (Belgium)
- Developed as part of a MSc thesis under supervision of prof. dr. ir. Kristel Bernaerts and prof. dr. ir. Steffen Waldherr
- Last updated: 14th April 2021
Theoretical basis
Sequential scanning algorithm based on pFBA, including MOMA extension to deal with genetically perturbed networks
- Scanning algorithm: Alper et al., Met. Eng., doi: https://doi.org/10.1016/j.ymben.2004.12.003
- MOMA integration: Wang et al., Biochem Eng J, doi: https://doi.org/10.1016/j.bej.2017.03.017
Contents
- The master function (
sequentialScanner.m) - The worker function (
sequentialScannerWorker.m) - Auxiliary function altering the model flux bounds for a gene upregulation (
upregulateModelGenes.m) - Auxiliary functions interconverting consensus gene names and systematic gene names (
GetGeneNameFromSysGeneName.m&GetSysGeneNameFromGeneName) MOMA.mfrom the COBRA Toolbox, with added bypass for wild-type solutions- An example overhead script calling the main function with appropiate parameters (
sequentialScannerOverhead.m): optimise ammonium efflux - The COBRA model necessary for this example script (
model100.mat): Bacillus subtilis in protein hydrolysate medium (genome-scale metabolic model iYO844 with custom-made GECKO extension)
Required software
- MATLAB 2019b or higher
- COBRA toolbox 3.0 for MATLAB
- IBM Cplex optimiser for MATLAB (or Gurobi Optimiser, or any other software capable of solving quadratic optimisation problems)
Other remarks
- This code currently uses a separate
geneNamesfield. If your metabolic model happens to contain a completegenesfield, just replace the appropiate field accessions.