Adaptive bi-level programming for optimal gene knockouts for targeted overproduction under phenotypic constraints

Area of Science:

• Metabolic Engineering
• Computational Biology
• Systems Biology

Background:

• Flux Balance Analysis (FBA) is a standard for steady-state metabolic analysis.
• Current gene knockout optimization often assumes maximal cell growth, which may not reflect reality.
• Minimization of Metabolic Adjustment (MOMA) better approximates knockout mutant phenotypes.

Purpose of the Study:

• To develop a new computational framework, MOMAKnock, for identifying gene knockouts.
• To improve targeted biochemical overproduction by incorporating MOMA.
• To derive robust gene knockout strategies under MOMA flux distribution.

Main Methods:

• A bi-level optimization framework is proposed, with chemical production as the primary objective.
• The inner problem constrains metabolic flux using the MOMA assumption.
• A novel adaptive piecewise linearization algorithm solves the resulting bi-level integer quadratic programming problem.

Main Results:

• MOMAKnock was tested on E. coli metabolic networks (core and iAF1260).
• Gene knockout strategies were compared to those derived from OptKnock.
• Preliminary results indicate improved targeted production and more robust strategies with MOMAKnock.

Conclusions:

• MOMAKnock offers a novel approach for metabolic engineering optimization.
• The MOMA assumption leads to improved knockout strategies for biochemical overproduction.
• This framework enhances the prediction of effective gene knockouts.