Regulatory on/off minimization of metabolic flux changes after genetic perturbations

T. Shlomi, O. Berkman, E. Ruppin
2005 Proceedings of the National Academy of Sciences of the United States of America  
Predicting the metabolic state of an organism after a gene knockout is a challenging task, because the regulatory system governs a series of transient metabolic changes that converge to a steadystate condition. Regulatory on͞off minimization (ROOM) is a constraint-based algorithm for predicting the metabolic steady state after gene knockouts. It aims to minimize the number of significant flux changes (hence on͞off) with respect to the wild type. ROOM is shown to accurately predict steady-state
more » ... etabolic fluxes that maintain flux linearity, in agreement with experimental flux measurements, and to correctly identify short alternative pathways used for rerouting metabolic flux in response to gene knockouts. ROOM's growth rate and flux predictions are compared with previously suggested algorithms, minimization of metabolic adjustment, and flux balance analysis (FBA). We find that minimization of metabolic adjustment provides accurate predictions for the initial transient growth rates observed during the early postperturbation state, whereas ROOM and FBA more successfully predict final higher steady-state growth rates. Although FBA explicitly maximizes the growth rate, ROOM does not, and only implicitly favors flux distributions having high growth rates. This indicates that, even though the cell has not evolved to cope with specific mutations, regulatory mechanisms aiming to minimize flux changes after genetic perturbations may indeed work to this effect. Further work is needed to identify metrics that characterize the complete trajectory from the initial to the final metabolic steady states after genetic perturbations. This paper was submitted directly (Track II) to the PNAS office. Abbreviations: FBA, flux balance analysis; MOMA, minimization of metabolic adjustment; ROOM, regulatory on͞off minimization; LP, linear programming; MILP, mixed-integer LP. ‡
doi:10.1073/pnas.0406346102 pmid:15897462 pmcid:PMC1140402 fatcat:srpulbcfarct7pefjod25yjm4m