Figure 1.
A metabolic network with one blocked reaction (A↔B).
Note that A appears with stoichiometric coefficient 2 in the boundary reaction →2A.
Table 1.
Summary of the main characteristics of GIMME [8], MBA [10], iMAT [35], mCADRE [26], INIT [13], and fastcore (this paper) reconstruction algorithms.
Figure 2.
Flowchart of the overall pipeline for generating consistent context-specific models.
Table 2.
Comparing fastcc to fastFVA [30] and CMC [10] on four input models.
Table 3.
Comparing fastcore to MBA [10] on liver model reconstruction from c-Recon1.
Figure 3.
Comparing fastcore to an exact MILP solver on a small E. coli model [38].
Shown are mean values of sizes of reconstructed models (over 50 repetitions for each core set; standard deviations were small and are omitted to avoid clutter) as a function of the size of the core set. fastcore computes near-optimal reconstructions, which improve with the size of the core set.
Figure 4.
Mean urea/glutamine ratio in the extended liver model obtained by fastcore.
Healthy (normal homozygote), partial (heterozygote) and full knock-out cases. See text for details.