Enumeration of minimal stoichiometric precursor sets in metabolic networks

Table 1 Our MILP approach (sasita) versus the combinatorial one

Strain	#Compounds/#reactions	#Sources/#targets	\(t_{sasita}\) (s)	\(t_{Combi}\) (s)
S. muelleri	76/64	9/Pyruvate	<1	<1
C. ruddii	128/126	45/Pyruvate	<1	1
B. aphidicola	282/245	91/l-Histidine	2	2
E. coli core^a	72/126	14/Biomass core	<1	42
E. coli core	78/126	14/Biomass core	<1	*
E. coli CFT073	1911/2949	26/Biomass core	12	*
E. coli EDL993	1895/2943	25/Biomass core	13	*
E. coli K-12	1806/2854	25/Biomass core	12	*
E. coli Sakai	1895/2942	25/Biomass core	12	*

^a Filtered network (see text). The MILP approach (sasita) and the combinatorial one (combi) are applied to several metabolic networks. For each instance, we provide the size of the network (number of compounds and reactions), the number of source and target compounds, and the time spent by both approaches (\(t_{sasita},\ t_{Combi}\)). One asterisk means that the combinatorial approach could not finish within the time limit

ISSN: 1748-7188